Company, ownership and the role

Proven rule. Every body claim traces to a source. Inferred or single-sourced-shaky claims live in section 4.

Kicker. MEDvidi is not a venture-backed startup with an absent product leader; it is a founder-controlled, cash-financed, two-entity American telehealth business whose engineering organisation was originally built end-to-end by a Belarusian software shop, Smart IT, whose CEO is MEDvidi's co-founder and titular CTO — and whose recruiters are running this search. The company took one small outside cheque (US$1.0m from AIM-listed TMT Investments in September 2021, carried at US$2.56m in mid-2024) and nothing meaningful since, while claiming to have reached roughly US$30m of annual revenue by May 2026. It had a CPO once; that seat has been publicly vacant since somewhere between November 2024 and March 2025, and the screening questionnaire behind the current VP of Product requisition was authored on 16 October 2025 — so this search is about ten months old, not ten days. Most importantly, the version of the posting Ilia was given is not the company's own version. MEDvidi's own careers page lists the role as a Contract, B2B, Poland / Remote-EU engagement with 22 vacation days, health insurance after probation, Preply English lessons — and no mention of equity at all. The word "equity" appears only in the recruiter-authored Notion copy. That single discrepancy is the most important thing in this Area, and it is checkable in one question.

Source legend

1. What we found

1.1 The corporate reality: two entities, one of which is not allowed to practise medicine

MEDvidi is not one company. Its own Member Terms & Conditions of Use open by naming two: "MEDVidi Health P.C. ("MEDvidi P.C"), a California professional corporation, through its medical director, and MEDvidi Inc. ("MEDvidi"), a Delaware corporation, (collectively, "MEDvidi Health") have entered a certain management services agreement" [LEGAL]. The Terms of Use page repeats the Delaware fact independently: "MEDvidi Inc., a Delaware corporation doing business as MEDVidi, owns and manages the Website" [LEGAL].

The division of labour is spelled out with unusual candour. "MEDvidi provides management and related services to MEDvidi P.C. … All of MEDvidi's services are non-medical and include, for example, intake assistance, scheduling assistance, software services, site maintenance, and technological support." And: "MEDvidi P.C. alone controls and supervises the Medical Services and clinical decisions reached by Practitioners" [LEGAL].

This is the standard American managed-services-organisation ("MSO" / "friendly PC") structure that corporate-practice-of-medicine rules force on any non-physician-owned clinical business. Three consequences are stated on the page rather than inferred:

• Fee-splitting firewall. In capitals: "ALL FEES, IF ANY, PAID TO MEDVIDI ARE SOLELY BASED ON THE FAIR MARKET VALUE OF ANY SERVICES THAT ARE PROVIDED TO YOU … FEES PAID BY YOU TO MEDVIDI ARE NOT BASED ON A PERCENTAGE OF PROFESSIONAL FEES RENDERED BY MEDVIDI P.C. OR THE VOLUME OR VALUE OF MEDVIDI P.C. SERVICES" [LEGAL]. The tech company is contractually forbidden from being paid per clinical dollar.
• PHI belongs to the P.C., not to the platform. "MEDvidi is a business associate of MEDvidi P.C. as such term is defined under HIPAA" [LEGAL]. The HIPAA Notice of Privacy Practices is issued in the name of "MEDVID HEALTH P.C." — the covered entity is the professional corporation [LEGAL].
• Clinicians are contractors of the P.C., not of the platform. "All Practitioners featured on the Sites are independent contractors of MEDvidi P.C. The Practitioners are independently insured, and any issue, warranty, or claim by a Member must be directly addressed to the respective Practitioner" [LEGAL].

Governing law and dispute forum: "The laws of the state of California … govern all adversarial proceedings", with venue in California state and federal courts, and binding arbitration administered by the American Health Lawyers Association under AHLA rules [LEGAL]. AHLA is the healthcare bar's own arbitration service — a drafting choice made by a healthcare-regulatory lawyer, not by a generic startup template mill.

One governance detail worth carrying into the interview: the HIPAA Notice names the Privacy Officer as "Oleg Gorbylev, COO" [LEGAL], and Gorbylev is listed on MEDvidi's own About page as COO of the group [ARC]. The privacy officer of the clinical covered entity is therefore the chief operating officer of the management company. That is common in this structure and not per se improper, but it is the seam where "clinical independence" is asserted and operational reality lives.

1.2 Address, phone numbers, and what they do and do not tell you

Every legal page and the footer give one address: 4010 Moorpark Ave., Suite 114, San Jose, California 95117 [LEGAL] [WEB]. LinkedIn's public company page lists headquarters as "San Jose, California", type "Privately Held", founded 2019 [LI]. The customer-support number in the footer is (504) 414-5095 — a New Orleans, Louisiana area code [WEB] — and a second number, 415-966-0848 (San Francisco), appears inside the Terms for security incidents, alongside team@medvidi.com, privacy@medvidi.com, help@medvidi.com and info@medvidi.com [LEGAL].

The Louisiana area code is a purchased number, not a Louisiana operation: Louisiana is not in the 35-state service list, and the corporate address, governing law and P.C. are all Californian. Treat the (504) number as telephony, not geography. (It does, though, appear in a suggestive pattern: MEDvidi ran a "Telehealth Mental Health Physician Remote (Louisiana)" requisition in December 2025 and again in July 2026 [ARC] — see §1.9.)

1.3 Founders: a healthcare operator and a software-shop owner

MEDvidi's own About page, in the version live from late 2023 until at least 3 January 2026, read: "Founded in 2019, MEDvidi emerged from offline clinics in San Francisco and Miami. Co-founders Vasili Razhnou and Alex Kulitski combined their expertise in healthcare services and technology to develop a platform where people across the US can get high-quality mental health support at an affordable cost" [ARC].

The published leadership team, as of the 3 January 2026 capture [ARC]:

Razhnou is the hiring manager. His public professional record: Belarusian State University; described as a serial founder with 15+ years in healthcare and business who has built five technology startups, with a background in business development, marketing strategy and IT consulting; entered healthcare in 2008 at a clinic still running on paper [PRESS-2026] [DIR]. MEDvidi's own timeline entry for 2019 is: "Vasili Razhnou founded an online mental health center MEDvidi after establishing traditional clinics in San Francisco and Miami" [WEB] — i.e. the founder is a clinic operator who went online, not a technologist. That matters for how he will grade a product candidate: he will test whether you understand the clinic's constraints, not whether you can recite AI vocabulary.

Alex Kulitski is the structural fact. He is Founder and CEO of Smart IT since October 2011 and Co-Founder and CTO of MEDvidi since September 2018 [DIR]; the 2021 joint press release describes him in exactly those dual terms — "Alex Kulitski, Smart IT CEO, Co-founder and CTO at MEDvidi" [PRESS-2021]; Smart IT's own newsroom confirms "Smart IT CEO and founder, Alex Kulitski" [SIT]. Public directory records place him in Dubai [DIR]. Before Smart IT he was co-founder and CTO of Bycard, described as Belarus's largest online ticket-booking service, 2012–2015 [DIR].

Note the asymmetry: outside MEDvidi he is described as MEDvidi's CTO; on MEDvidi's own site he has never been given the CTO title — he was "Co-founder / R&D Team Lead" in 2024 and plain "Co-founder" by 2026 [ARC]. There is no CTO in MEDvidi's published leadership list at any point. The posting instead puts a VP of Engineering under the VP of Product [JD] [JD-SITE]. Who actually decides engineering direction — the co-founder who built the platform and is called its CTO everywhere else, or the incoming VP of Product — is not answerable from public sources. It is the single most important unknown in this Area (§4.1).

1.4 Smart IT: not a parent, not a passive vendor — the shop that built the product and still staffs the search

Smart IT describes itself as "a small software development company made up of business and software development professionals. We make the most of modern technologies to power the ideas of our clients. Oh, and we run a few projects of our own." [SIT]. Its self-reported quick facts — undated, on a site whose footer says © 2020 — are: 14 years in operation, 60+ employees and growing, 46 clients worldwide, 2M+ users of products it helped develop, and a "10:33 girls-to-boys ratio" [SIT]. Its published senior team is three people, first names only: Alex (Founder, CEO), Egor (CTO), Denis (CBDO) [SIT]. Its "what we DON'T" list includes "hire junior developers" and "subcontract" [SIT]. Registered address on the site: 30 N Gould St Ste R, Sheridan, Wyoming 82801, with a +1 408 (San Jose) phone [SIT] — a well-known registered-agent address, i.e. a US shell for a company whose delivery centre is elsewhere. Case-study locations skew heavily to Minsk, Belarus, and the collaboration tooling listed (Skype, Viber, Worksection) is a 2015–2020 Eastern-European stack [SIT].

The relationship to MEDvidi is documented by both parties:

• The 2021 joint release: "On the initiative of MEDvidi, the Smart IT team has developed a telemedicine platform from scratch" [PRESS-2021].
• Smart IT's case study, in its own words: "Smart IT is responsible for the entire development of MEDvidi, a fully functional telemedicine platform. At the moment, the platform integrates and combines all medical services provided by our Client's proprietary online and offline clinics" [SIT-CASE].
• The scope it claims to have built: "a robust video and voice conferencing tool, custom CRM and EHR/EMR systems compliant with HIPPA requirements", e-prescribing, Twilio telephony, SendGrid notifications, payment-processor integrations, automated scheduling and automated return visits, and separate doctor and patient dashboards [SIT-CASE].
• Its claimed engagement metrics: 20 months of development, 11 specialists on the team, 18 technologies, 270% business revenue growth [SIT-CASE].
• The homepage card that links to this case study specifies the team shape: "5 software engineers, 2 QA engineers, 1 BA, 1 PM, 1 UX/UI specialist, CTO (Executive-as-a-service model)", San Francisco, HealthTech, 19 months, status active [SIT].

So Smart IT did not merely write code — it rented MEDvidi its CTO as a service, and that CTO is Smart IT's own CEO and MEDvidi's co-founder. Two roadmap items are listed as "in progress" in that case study and are worth remembering: "insurance claim management and pharmacy network integrations" [SIT-CASE].

On the venture-builder question the honest answer is no, but. Smart IT's own-product history amounts to two named ventures across fifteen years — Bycard (2012–2015) and MEDvidi (2018–) [DIR] [SIT]. This is not a portfolio of ten experiments where MEDvidi is one bet among many; MEDvidi appears to be the own-product bet of a services firm whose founder went all in on it. That is a materially better situation for an incoming VP of Product than a true venture studio would be — but it also means the co-founder's identity is fused with the platform he personally architected.

Finally, the recruiting signal. Both recruiters on the posting Ilia received use @smart-it.io addresses [JD]. MEDvidi simultaneously runs its own ATS under its own name [ATS], and posted its own "Talent Acquisition Specialist" and "Talent Acquisition Specialist (Healthcare)" requisitions in April 2026 [ARC] plus four "People Operations Specialist" requisitions in August 2026 [ATS]. So MEDvidi is building an in-house people function while its executive search still runs through Smart IT's staff. Whether Smart IT is also the contracting or employer-of-record vehicle for MEDvidi's Eastern-European engineers is unresolved (§4.2) — and it is the question that decides whether the 30+ engineers are the VP of Product's team or somebody else's account.

1.5 Ownership and funding: one small cheque, then cash

MEDvidi appears on the public portfolio page of TMT Investments PLC, a venture fund listed on London's AIM (ticker TMT), described there as an "online mental health solution that provides affordable virtual medication management and online mental care" [OWN]. That is a primary, verifiable ownership link: a publicly listed fund names MEDvidi as a holding.

Search-surfaced TMT disclosure gives the numbers: an initial US$1.0 million investment in "MedVidi Inc.", dated 27 September 2021, carried at US$2,560,000 as at 30 June 2024, with a note that MedVidi "received further validation of their progress by raising fresh capital at valuation levels that resulted in positive revaluations for TMT", and a further positive revaluation as at 31 December 2024 [OWN]. I could not open the underlying TMT half-year report to quote the row verbatim — no PDF text extraction was available in this environment — so treat the two figures as search-surfaced rather than document-verified (§4.5). The qualitative claim is the load-bearing part: there was a priced round after 2022.

Aggregator figures disagree on totals: one reports US$3.0m raised in a single seed round in September 2022; another reports US$1.83m; named investors across sources include Advantary, AltaIR Capital, The Garage Syndicate, Verras Capital, TMT Investments and Gaingels [DIR]. These are aggregator claims and should be treated as such. What survives triangulation is the shape, and the shape is what matters:

Outside capital raised, all sources, 2019-2026:   US$1.8m - US$3.0m  [DIR]
Of which TMT's initial cheque:                    US$1.0m (27.09.2021)  [OWN]
Revenue attributed to the company by May 2026:    ~US$30m ARR  [PRESS-2026]
Ratio of revenue to lifetime outside capital:     30 / 2.5 ~= 12x   [REASON]

A company at roughly US$30m of revenue that has raised about US$2–3m in its life is, for all practical purposes, bootstrapped after a friends-and-angels round. Note what is absent: no Series A, no B, no institutional lead, no press cycle. In a category where competitors raised nine-figure rounds, that absence is itself the finding.

For scale calibration, TMT's own executive director said in a June 2025 RNS that TMT's position in Bolt alone exceeded TMT's entire market capitalisation, "which effectively means that TMT's portfolio of around 50 other investments is assigned negative value" [OWN]. MEDvidi is one of that ~50. No fund on this cap table has the incentive or the weight to force a growth strategy on the founder.

Three implications a VP of Product should hold:

1. Decisions get made by one person for cash reasons. There is no board seat that needs a growth narrative, and no runway clock. The CEO optimises contribution margin and payback, not ARR-at-any-cost. A product proposal that costs money before it makes money must be argued in payback months, not in strategic narrative.
2. Equity here is not a lottery ticket. With no institutional lead, no priced-round cadence, and a founder who controls the company and does not need to sell, the exit forcing-function is weak. Any equity component should be interrogated on instrument (options? phantom? which entity — the Delaware Inc.?), strike, most recent 409A or priced valuation, vesting, and what happens on a contract termination rather than an employment termination.
3. The company has already survived its own growth. Going from 12 states to 36 and from 35k to 120k annual appointments on US$2–3m of outside money [WEB] means the operating model produces cash. That is a genuinely strong hand, and it is the honest reason the CEO can offer a C-suite-sized scope on a contractor's terms: he does not have to compete with funded companies for talent on cash, only on scope.

1.6 Headcount and the shape of the company

The self-reported people numbers are inconsistent across the company's own pages (§3.2) and should not be quoted to the CEO as fact. What is reliable is the composition, and that is computable from the posting itself:

Product & engineering organisation, per the canonical posting [JD-SITE]:
  1  Director of Product
  8  Product Managers
  1  Director of UX & Design
  3  Designers
  1  Director of Data & Analytics
  6  Data Analysts
  1  Data Engineer
  1  Lead PMM
  1  VP of Engineering
 30+ Engineers & QAs
 ---
 53  people minimum, + the VP of Product = 54            [REASON]

Direct reports to the VP of Product: 5
  (Dir Product, Dir UX & Design, Dir Data & Analytics, Lead PMM, VP Engineering)  [REASON]

Ratios:
  Engineers & QAs per PM        = 30 / 8   = 3.75    [REASON]
  Designers per PM              =  3 / 8   = 0.38    [REASON]
  Analysts (incl. dir + eng) per PM = 8 / 8 = 1.00   [REASON]
  Analysts per data engineer    =  7 / 1   = 7.0     [REASON]
  Product+eng share of 193 LinkedIn profiles = 54/193 = 28%   [REASON]
  Revenue per product+eng head  = $30m / 54 = ~$555k [REASON, uses the $30m figure]

Read those ratios rather than the headcount. Four things fall out:

• 3.75 engineers per PM is not a delivery organisation, it is a discovery organisation. Eight PMs cannot each run a squad on 30 engineers. Either several PMs own workflows rather than teams, or the PM population has drifted into analyst/coordination work. The company simultaneously has 6 data analysts and posts "Product Analyst (Clinical Guidance)" and "AI Product Analyst — Voice AI Agents" requisitions in the Product Management and Engineering departments respectively [ATS] — more analysis capacity being added to an already analyst-saturated org.
• One analyst per PM, and one data engineer for seven analysts. This is the signature of a company that answers questions by hand. Seven people consuming data and one building the pipes means dashboards are probably bespoke, definitions probably differ between analysts, and any "segment → root cause → fix" instruction in the posting [JD] is currently executed as bespoke SQL per question. That is precisely why the posting demands that diagnostic method by name.
• 0.38 designers per PM across two tracks. Three designers for a consumer funnel, a patient portal, a clinician workspace and an AI clinic is starvation-level. Consistent with this, the company posted a Head of Product Design requisition in July 2026 [ARC] and currently keeps a "Join MEDvidi Design Talent Pool" evergreen req open [ATS]. The design director named in the posting was hired between July and August 2026 — a brand-new peer.
• A dense director layer relative to span. One Director of Product over 8 PMs, one design director over 3 designers, one data director over 7. The company has built a management tier before it has built the pipeline and craft capacity underneath it.

1.7 "Two Directors of Product" is a recruiter's paraphrase, and the org is different from what Ilia was told

This is a hard, checkable discrepancy. The two texts do not match.

Everything else — responsibilities, all twelve requirements, the three shipped AI initiatives, the "reports directly to the CEO" line, the two named tracks — is word-for-word identical [JD] [JD-SITE]. So the divergences are deliberate edits, not drift.

Two of them change the job:

1. The design director is not a second product director. "2 Directors of Product" reads as a redundant management layer over 8 PMs; the real structure is one product director plus a brand-new design director. That is a materially better org than the paraphrase suggests, and it means the VP inherits a design function that has just been given a leader — a peer with something to prove, not a vacancy.
2. The employment relationship is a B2B contract with a probation period, and the company's own posting does not offer equity. The benefits listed are a standard Warsaw/Belgrade contractor package. The word "equity" exists only in the recruiter's block. Either the recruiters are authorised to offer something the company's public posting omits, or the Notion page overstates. Both possibilities are live; only the CEO can settle it.

1.8 The role has been open, in some form, since October 2025 — and there was a CPO before

Three independent dating methods converge.

Method 1 — the leadership page. Julia Guzman is listed as CPO on MEDvidi's About page in every capture from 4 January 2024 through 8 November 2024, and is absent from the 19 March 2025 capture and every capture after [ARC]. Wayback has no captures between those two dates, so the vacancy began between 2024-11-08 and 2025-03-19. Also in that window the CMO's title changed from "Chief Medical Director (CMD)" to "Chief Medical Officer (CMO)" (between 16 March and 2 June 2024) [ARC].

Method 2 — the ATS questionnaire. Comeet stores each screening question with a millisecond-epoch uid set when the question is authored. The VP of Product requisition's four questions carry uids 1760633980457–1760634096807, which decode to 16 October 2025, 16:59–17:01 UTC [ATS]. No other live requisition shares that date: the engineering and product IC reqs date from February–August 2026, the design and talent-pool reqs from June 2026, the clinical template from 2 February 2025 [ATS]. There was no prior VP of Product to clone a questionnaire from. The most economical explanation is that the requisition itself was opened on 16 October 2025 and has been running, publicly or privately, ever since — about ten months as of this Inquiry.

Method 3 — the public board. The VP of Product role was not listed on medvidi.com/careers in the 4 July 2026 capture, and is listed on 17–18 August 2026 [ARC] [ATS]. Wayback has no captures in between. So the public posting is at most six weeks old.

Put together, the picture is not "stale posting reposted for a year" and not "brand-new opening" either. It is: a product-leadership seat vacated around the turn of 2024/25, a requisition opened in October 2025, and a public relaunch of that requisition in roughly late July or early August 2026 — the relaunch scoped much larger than the seat it replaces, because it now swallows Engineering, Design, Data and PMM.

The four screening questions themselves are informative [ATS]:

1. "Do you have 10+ years of experience in product management?"
2. "Please briefly describe your experience building AI-first products using RAG systems, agentic AI, and AI evals" — the only free-text question, i.e. the actual filter.
3. "Do you speak Russian fluently?"
4. "Are you comfortable working in CET or GMT+4 time zones?"

Question 4 is the one nobody expects. The role is not on US hours. CET covers the Poland / Serbia / Spain / Portugal engineering belt; GMT+4 covers Armenia, Georgia, the UAE — and the co-founder/CTO's public location is Dubai [DIR]. The product leader of a US-only, US-clinician, US-patient business is expected to sit 6–11 hours ahead of the market. Every clinician, every patient and every regulator is on US time; every engineer, designer, analyst and executive is not.

1.9 What the hiring board says the roadmap actually is

MEDvidi's Comeet board carried 59 live requisitions on 2026-08-18, resolving to 22 distinct roles once geographic duplicates are collapsed [ATS]:

The distinct non-clinical roles open right now: AI Product Analyst — Voice AI Agents (4 geos; its own URL slug reads ai-product-analyst-ai-receptionist-team-voice-ai-agents, naming an AI Receptionist team), Engineering Manager (4 geos), Staff DevOps Engineer (3 geos), Product Analyst (Clinical Guidance) (3 geos), People Operations Specialist (4 geos), VP of Product (3 geos), plus evergreen Product / Design / Engineering talent pools [ATS].

The Wayback history of the same board is the better artefact, because it shows what the company decided to do quarter by quarter [ARC]:

Two threads deserve to be pulled in the interview.

Thread one: insurance. MEDvidi's public position is cash-pay, no insurance [WEB]. Yet (a) Smart IT's case study lists "insurance claim management" as in progress [SIT-CASE]; (b) the Member Terms & Conditions say "You may be asked to provide your insurance information … By providing your insurance information, you authorize MEDvidi Health to submit claims and bill Medical Services to your insurer on your behalf" [LEGAL]; and (c) the company posted a Director, Revenue Cycle Management in July 2026 [ARC]. Three independent artefacts, one direction. Either an insurance-billed line already exists somewhere in the business, or it is being built. See §3.1.

Thread two: clinician supply is the standing constraint. Of 59 live requisitions, 35 are clinical, covering 33 of the 35 states MEDvidi says it serves — every one except North Dakota and New Mexico [ATS] [REASON]:

States MEDvidi says it serves:                       35   [WEB]
States with a live clinician requisition (2026-08):  33   [ATS]
Coverage:                                            33/35 = 94%   [REASON]
Clinical share of all live requisitions:             35/59 = 59%   [REASON]

A company recruiting prescribers in 94% of its markets at once is either running a permanent 1099 top-of-funnel or is capacity-short everywhere. The CEO's own framing supports the second reading: "within three months of joining us, most providers are 80% booked with follow-up patients" [PRESS-2026]. New clinicians fill up in a quarter. And the utilisation arithmetic is consistent with a part-time contractor pool rather than a salaried staff:

Annual visits (company's own figure):         120,000   [WEB]
Providers (company's own 2025 figure):            117   [WEB]
Visits per provider per year: 120,000 / 117 =   1,026   [REASON]
Per 48-week year:            1,026 / 48     =    21.4 visits/week   [REASON]
At 15-min follow-ups + 30-min intakes, ~21 visits/week is roughly
a half-day-to-one-day-a-week engagement per clinician.   [REASON]

The careers page reinforces it: "work on your own schedule", "Flexible hours", "Work part-time or shift to W2", "MEDvidi sponsors your medical and DEA licenses", referral bonuses, and an on-page earnings calculator with a 5–50 hours-per-week slider [WEB]. The clinical workforce is a marketplace-like contractor pool that MEDvidi pays to license.

1.10 One clinical entity, three consumer brands

The About page timeline records, for H1 2023: "MEDvidi unites with EZCare Clinic and Mango Clinic", alongside "300 people, 14 states, 36,000 appointments in half a year, +75% revenue growth, doubled the number of providers, 50% of healthcare providers are MD" [ARC]. That consolidation is verifiable today from the sibling sites themselves:

• ezcareclinic.com 301-redirects to ezcareclinic.io, which is a live, separately branded funnel with its own Google Tag Manager container (GTM-5G78WWS, versus MEDvidi's GTM-T7TBQ8C) — and whose Terms & Conditions are MEDvidi's, naming "MEDVidi Health P.C.", "MEDvidi Inc.", the fair-market-value fee language and the 4010 Moorpark Ave address [BRAND].
• mangoclinic.com 301-redirects to medvidi.com, and its /about-us/ path still serves MEDvidi's About page with "© MEDvidi, Inc." and Razhnou's founding story [BRAND].

So: Mango has been absorbed into the MEDvidi brand; EZCare is still live as a second front door on the same legal and clinical entity. And it is not a clone. EZCare's FAQ lists 20 prescribing states — including New Jersey and Oklahoma, which are not among MEDvidi's 35 — and sells "60-minute follow-up therapy sessions" with a "personal licensed therapist" alongside 15-minute prescriber follow-ups [BRAND]. Different state pool, different service mix, different analytics, same P.C.

That is a material fact about the job that appears nowhere in either version of the posting. The product surface being inherited is at least two consumer brands with different footprints on one clinical back end — which multiplies every funnel, every state-eligibility rule, every pricing table and every analytics definition by two.

1.11 Employee-side signals — thin, and to be used as texture only

Glassdoor shows 4.0/5 from 17 reviews, 73% would recommend, sub-ratings of 4.2 work-life balance, 4.2 culture and values, 3.9 career opportunities, and 3.4 compensation and benefits — the lowest dimension. Interview experience is 60% positive with an average difficulty of 2.8/5, and one review describes the hiring flow as "screen → panel → BG/refs → offer" with quick decisions [REV].

Recurring pros: fully remote with real flexibility, friendly and helpful atmosphere, meaningful work, offline company events, "management listens to feedback and actually makes changes" [REV]. Recurring cons: "things move fast and require quick adaptation, with some processes still being improved as the company grows"; "upper management micromanages everything"; and — from a provider-side perspective — "language barriers among support staff who have difficulty understanding provider needs" [REV].

n≈17, self-selected, no way to weight by function or tenure. This is not a sample. But the last con corroborates something structural rather than anecdotal: US clinicians on one side, a support and administrative layer hired in Pakistan, Turkey and Tbilisi on the other [ARC], mediating clinical workflow. The lowest Glassdoor dimension being compensation, in a company that hires almost everyone on Eastern-European B2B contracts, is also not a surprise.

1.12 What "competitive compensation" plausibly means here

Reason it out rather than guess it. The requisition is Contract, located Poland / Remote EU / Remote, under an explicitly stated B2B model with a probation period, 22 vacation days and Polish-market perks [JD-SITE] [ATS]. It will therefore be benchmarked against the Warsaw/Belgrade contractor market for a director-or-above technology leader, not against US VP Product comp.

Poland B2B benchmark, 2026, director / C-level average:   31,659 PLN / month   [COMP-PL]
VP-equivalent / technical co-founder band:                32,000-50,000+ PLN / month   [COMP-PL]
Assumption: FX ~= 3.6 PLN per USD (mid-2026 order of magnitude; NOT a sourced rate)

  31,659 PLN/mo x 12 = 380k PLN/yr  ->  ~US$106k/yr
  50,000  PLN/mo x 12 = 600k PLN/yr  ->  ~US$167k/yr
  70,000  PLN/mo x 12 = 840k PLN/yr  ->  ~US$233k/yr  (top-of-band stretch)

These are NET INVOICE amounts under Polish B2B: the contractor pays their own
tax and social contributions out of them.  [COMP-PL]

For comparison, a US-employed VP Product owning PM+Eng+Design+Data+PMM at a
~$30m-revenue consumer health company would customarily be in the
US$250-350k base range plus meaningful equity.  [REASON, no US benchmark pulled]

Implied gap: roughly 1.5x-3x on cash, before any equity.   [REASON]

This is an estimate. I did not pull a US VP-Product benchmark for this Area and the FX rate is an assumption, not a source. But the direction is not in doubt, and it explains the posting's rhetoric. "A scope typically reserved for C-suite executives at much larger companies" [JD] [JD-SITE] is doing real work in that sentence: the company is offering title and scope in place of US cash and liquid equity. That is a legitimate trade — for the right candidate it is the fastest route to a CPO-shaped resume line and a genuinely broad mandate — but it should be named as the trade it is, not received as flattery.

On the equity specifically: in a founder-controlled Delaware C-corp with US$1–3m of lifetime outside capital, one small AIM-fund holder among ~50 positions, no institutional lead and no exit clock [OWN] [DIR], an option grant to a contractor is a low-probability, illiquid instrument. Non-employee contractors typically cannot receive incentive stock options at all; they receive NSOs or a phantom/appreciation-rights arrangement, which changes the tax and the mechanics. Ask which instrument, in which entity, at what valuation, and what happens to it if the B2B contract is terminated rather than employment.

1.13 The honest read on the role

Weighing the evidence, four hypotheses and what supports each.

H1 — Genuine platform of trust, deliberately over-scoped. Support: the CEO is a clinic operator, not a technologist [WEB] [PRESS-2026]; the CPO seat has been empty for ~18 months and the CEO has evidently been the de facto product owner since; the org has been built out underneath (directors of product, data and design all in place) while the top seat stayed empty; the CEO's public interview is entirely about product mechanics — charting time, follow-up automation, chart review, receptionist agents [PRESS-2026]. A founder who talks like that wants a product partner, not a project manager. Weight: strong.

H2 — Title-and-scope arbitrage in place of cash and equity. Support: Contract/B2B employment type; benefits denominated in Polish-market terms; no equity in the company's own posting; the "C-suite at much larger companies" line; Glassdoor's lowest score is compensation [JD-SITE] [ATS] [REV]. Weight: strong. Not mutually exclusive with H1.

H3 — Backfill after a departure. Support: Julia Guzman held CPO until at least November 2024 and was gone by March 2025 [ARC]; the requisition dates to October 2025 [ATS]. Against: the new role is much larger than CPO-over-product — it absorbs Engineering, Design, Data and PMM. Weight: moderate. It is a backfill that has been upgraded.

H4 — The founder wants to stop running product, but not stop deciding. Support: the entire leadership section was deleted from the About page between 3 January and 6 June 2026 [ARC]; the co-founder is called MEDvidi's CTO everywhere except on MEDvidi's own site [DIR] [SIT] [ARC]; the posting hands the VP a VP of Engineering while the co-founder who personally architected the platform remains in place; Glassdoor's sharpest criticism is "upper management micromanages everything" [REV]; and the role is a contract, which is the easiest senior relationship to unwind. Weight: unresolved, and it is the risk that matters.

My read, labelled as a read: H1 and H2 are both true, and H4 is the thing to test. The mandate is real — there is genuinely no one else doing this job, the AI surface is genuinely shipped rather than aspirational, and the business genuinely generates cash to fund product work. But it is priced as an Eastern-European contract, and the boundary between the incoming VP of Product and the co-founder/CTO who built the platform is undefined in every public document. That boundary is the whole job.

2. Capability / object table

3. Reconciliation notes

Reconciliation note — insurance. [WEB] says MEDvidi is cash-pay and accepts no insurance, and that is the consumer-facing truth confirmed across the whole marketing site. But [LEGAL] (Member Terms §3(b)) says "You may be asked to provide your insurance information … you authorize MEDvidi Health to submit claims and bill Medical Services to your insurer on your behalf"; [SIT-CASE] lists "insurance claim management" as an in-progress workstream; and [ARC] shows a Director, Revenue Cycle Management requisition posted in July 2026. Resolved as: the current retail offer is cash-pay ([WEB] wins for today), and an insurance-billed capability is either latent, partial, or actively being built ([LEGAL] + [SIT-CASE] + [ARC] win for direction). Not resolved: whether any revenue is billed to payers today. Handed to A3.

Reconciliation note — org composition. [JD] says "2 Directors of Product"; [JD-SITE] says "1 Director of Product … 1 Director of UX & Design". Resolved in favour of [JD-SITE], because it is the company's own posting on its own ATS, it is internally consistent with the July 2026 "Head of Product Design" requisition in [ARC], and a recruiter compressing two different directorships into "2 Directors of Product" is a far likelier error than a company mis-describing its own reporting lines on its own careers page.

Reconciliation note — equity. [JD] promises "Competitive compensation and equity aligned with impact". [JD-SITE] offers "a competitive compensation package" plus Polish contractor benefits and says nothing about equity anywhere. Not resolved, deliberately: both texts are authentic, and the difference could be either recruiter licence or a genuine negotiable that the company chose not to publish. This is an interview question (§6), not a finding.

Reconciliation note — headcount. The company's own numbers do not agree with themselves. The current About page timeline reads "Growth from 50 to 200+ people" [WEB], while the About page live until June 2026 read "grown from 50 to 300+ dedicated team members" and "300 people" at H1 2023 [ARC]. LinkedIn's band is 201–500 with 193 associated profiles [LI]; a third-party directory says 51–200 [DIR]. Resolved as: no reliable total headcount exists in public sources. Use the composition (§1.6) and the 193 LinkedIn figure as the only defensible anchor, and never quote a total to the CEO as if it were known.

Reconciliation note — states. The current About page says "Open states: 36" and "Growth from 12 to 36 states in 3 years" [WEB], while the service pages list 35 states (Inquiry spine, [WEB]), and EZCare lists 20 states including New Jersey and Oklahoma, which appear in neither [BRAND]. Resolved as: 35 is the operative, checkable list; 36 is a marketing round-up on a page whose timeline panels are also internally scrambled (the 2024 panel shows fewer states and providers than the 2023 panel [WEB]). The EZCare discrepancy is not reconciled and is parked (§4.7).

Reconciliation note — funding total. [DIR] sources give US$3.0m (one seed, Sept 2022) and US$1.83m for the same company. [OWN] independently establishes a US$1.0m TMT cheque dated 27.09.2021 — a year earlier than the "September 2022 seed", which suggests the aggregators have mis-dated or merged rounds. Resolved as: the total is somewhere in the US$1.8m–3.0m range and the exact figure is not knowable publicly. The [OWN] datapoint is the only one with a listed-company audit trail behind it.

Reconciliation note — the $30m ARR figure. [PRESS-2026] carries "Under his leadership, the company reached $30M in ARR", but in the interview's introductory bio rather than as a CEO quotation, so its provenance is an editor summarising a provided bio. It also does not reconcile with the company's own volume and price figures: 120,000 annual visits [WEB] at the published $195 initial / $159 follow-up [WEB] implies roughly US$19–23m, i.e. $30m requires an average of $250 per visit. Not resolved. Flagged to A3 as the single most important number to triangulate, with the arithmetic below.

120,000 visits x $159 (all follow-ups)  = $19.08m   [REASON]
120,000 visits x $195 (all initials)    = $23.40m   [REASON]
$30m / 120,000 visits                   = $250 per visit   [REASON]
=> $30m ARR is unreachable at published visit prices on 120k visits.
   Either (a) visits are now materially above 120,000,
          (b) revenue includes non-visit lines (subscriptions, ESA letters,
              weight-loss programmes, affiliate commissions - the latter
              disclosed in [LEGAL] section 23), or
          (c) the $30m figure is forward-looking or overstated.

4. Open Questions / Parked

4.1 Who owns engineering — the VP of Product, or the co-founder who built the platform? Both versions of the posting place a VP of Engineering under the VP of Product [JD] [JD-SITE]. Alex Kulitski is described as MEDvidi's CTO by Crunchbase, by his own public LinkedIn headline, and by the 2021 joint press release [DIR] [PRESS-2021], but has never carried that title on MEDvidi's own site [ARC]. What would settle it: asking the CEO directly what Kulitski's role is today, whether the VP of Engineering reports to the VP of Product or to the co-founder, and who signs off architecture. There is no public document that can answer this.

4.2 Is Smart IT the employer-of-record or contracting vehicle for MEDvidi's non-clinical staff? MEDvidi hires directly under its own name on its own ATS, in Poland, Serbia, Spain, Portugal, Armenia, Georgia and elsewhere, on B2B contracts [ATS]. Its exec search is run by people with @smart-it.io addresses [JD]. Smart IT sells "Dedicated Development Team" and outstaffing [SIT]. What would settle it: asking which legal entity issues the contract for the VP role and for the engineers, and whether any of the 30+ engineers are Smart IT staff on an MSA rather than MEDvidi contractors. This determines whether the VP of Product can hire, fire and re-shape the engineering team at all.

4.3 Is Laura Purdy still Chief Medical Officer, and who is the P.C.'s medical director? She was listed as CMO in every About-page capture up to 3 January 2026 [ARC], is absent from the current site entirely, does not appear on /providers/ or among the 35 team profiles [WEB], and her public professional record shows a portfolio of fractional medical-director/advisor roles across "more than 50 telehealth providers", plus her own venture [DIR]. Separately, the Terms refer to "MEDvidi P.C. … through its medical director" without naming that person [LEGAL]. What would settle it: asking who the P.C.'s medical director is, whether the CMO role is full-time or fractional, and who owns clinical protocol change control. This matters because Chart Review AI encodes clinical protocol, and protocol ownership is not the VP of Product's to take.

4.4 Why was the entire leadership section deleted from the About page between January and June 2026? [ARC] shows it present on 2026-01-03 and gone on 2026-06-06, along with the EZCare/Mango consolidation narrative, the co-founder attribution, and the detailed history. Possible explanations range from a routine redesign to executive churn to a deliberate choice to stop publishing named executives. What would settle it: one question, and the reaction to it, in the interview.

4.5 Corporate registry records are unverified. California's bizfileonline.sos.ca.gov API is behind an Imperva/Incapsula WAF that refused programmatic access, and Delaware's entity search requires an interactive session. So the Delaware incorporation of MEDvidi Inc., the California registration of MEDVidi Health P.C., their formation dates, officers and registered agents are asserted only by the company's own legal pages [LEGAL]. TMT's per-holding figures (US$1.0m initial, US$2.56m carrying value) are likewise search-surfaced rather than quoted from the report, because no PDF text extraction was available in this environment. What would settle it: a manual lookup in the two registries, and opening TMT's 2024 half-year report and 2025 annual report by hand.

4.6 Are there other consumer brands on the same P.C. beyond MEDvidi, EZCare and Mango? Two were named in the H1-2023 consolidation [ARC] and both are accounted for [BRAND]. I did not enumerate the domain estate systematically — no reverse-WHOIS, no analytics-ID sweep across candidate domains — so a fourth brand could exist. What would settle it: a reverse lookup on the GTM container IDs and the shared legal text, or simply asking how many consumer brands the product org supports.

4.7 Why does EZCare list New Jersey and Oklahoma when MEDvidi lists neither? [BRAND] gives 20 EZCare prescribing states including NJ and OK; the MEDvidi spine list of 35 excludes both. Either EZCare's FAQ is stale, or the two brands genuinely draw on different licence pools. What would settle it: checking EZCare's booking flow for NJ/OK availability, or asking how state eligibility is modelled across brands.

4.8 Was the VP of Product requisition actually live between October 2025 and July 2026? The questionnaire timestamp says the req was authored 2025-10-16 [ATS]; the public board did not show it on 2026-07-04 [ARC]. Comeet question uids are client-side timestamps and would be carried along if a requisition were cloned — but there is no plausible donor requisition, since no prior VP of Product existed. What would settle it: asking how long they have been looking and how many candidates have reached final stage. The answer to "how long" is a good honesty test, because the October 2025 timestamp is checkable.

4.9 What is the actual compensation band, and is equity real? Section 1.12 is arithmetic over Polish market benchmarks with an assumed FX rate, not a sourced band for this role. No US VP-Product benchmark was pulled. What would settle it: the recruiters, who will have a band; and the CEO, on the equity instrument.

4.10 Who are the "advisors"? [JD] promises "Close partnership with the CEO, executive leadership, and advisors" — a phrase absent from the company's own posting [JD-SITE]. No advisory board is published anywhere. What would settle it: asking who the advisors are and what they advise on; in a cash-financed company with a fractional CMO, "advisors" often means the regulatory counsel and the medical director, which would be useful to know.

4.11 Where does the CEO physically sit, and where does the exec team meet? The GMT+4 option in the screening question [ATS] and the co-founder's Dubai location [DIR] suggest a Gulf or Caucasus centre of gravity, but I have no public source for the CEO's location. What would settle it: asking, plainly, where the leadership team is and how often it is in a room.

5. What this does NOT cover

• The product itself. Nothing here describes the Personal Portal, the clinician workspace, the funnel, or how AI Scribe / Chart Review AI / the Agentic AI Receptionist actually behave. A1 and A5 own that. This Area touches those products only where a hiring requisition or an org line reveals something about who builds them.
• Unit economics and the revenue model. I computed the $30m-versus-120k-visits contradiction because it surfaced from a company-org source and needed flagging, but I did not attempt CAC, LTV, contribution margin, retention or refund exposure. A3 owns that, and should treat §3's ARR arithmetic as an input.
• Regulation. The Compliance Director requisition of December 2025 and the Director of RCM requisition of July 2026 are reported here as hiring signals only. The DEA telemedicine rule, the 2026-12-31 expiry, state-by-state prescribing law, and enforcement precedent belong to A4.
• Clinical supply mechanics. I report the 33-of-35-states requisition density, the earnings-calculator model, DEA-licence sponsorship and the ~21-visits-per-provider-per-week arithmetic because they are org facts. Panel composition, licensure spread, collaborative- practice requirements and provider unit cost belong to A5.
• Competitive positioning. EZCare appears here as a corporate-structure fact, not as a market position. Who else takes this click belongs to A6.
• Any private-individual information. Everything about named people in this Finding comes from what MEDvidi, Smart IT, the press releases or those people's own public professional profiles publish about their professional roles. No personal-life, family, residential, financial or non-professional information was sought, found or recorded — and none should be. Where a name appears without a verified current title, it is parked rather than guessed.
• The 35 clinician and writer profiles on medvidi.com. Enumerated as a count from the team sitemap; individual credentials, licences and state coverage are A5's to analyse.
• Corporate registry primary records. Blocked (§4.5). Everything about entity formation is the company's own assertion.
• Whether Smart IT has other clients that matter. Its case-study list was read for the MEDvidi entry and for the shape of the firm; I did not assess its other 45 clients.

6. What this means for a VP of Product

The structural fact that governs everything else is that this is a two-entity business, and the product organisation sits inside the entity that is legally forbidden from practising medicine. MEDvidi Inc. sells software, scheduling, intake and support to MEDVidi Health P.C., at fair market value, explicitly not as a share of clinical revenue [LEGAL]. Every AI feature the company has shipped lives on the Inc. side of that wall and touches work that belongs to the P.C.: AI Scribe writes the note the clinician signs, Chart Review AI flags deviations from protocols the P.C.'s medical director owns, the AI Receptionist handles patient contact. A new VP of Product will be tempted to treat clinical protocol as product configuration. Do not. The correct posture is to own the tooling, evidence and evaluation around clinical decisions and to make protocol change a P.C.-governed process with a product-grade workflow — versioned, auditable, with an owner and a review cadence. Building that governance surface is a real, shippable product, and it is the single most defensible thing a new product leader can do in this company because it makes AI expansion safe rather than merely fast.

The second fact is that the money is the founder's, which changes how you sell work internally. US$1.8–3.0m of lifetime outside capital against roughly US$30m of revenue [DIR] [OWN] [PRESS-2026] means there is no board narrative to satisfy and no runway countdown — but also no tolerance for a roadmap whose payback is a story. Bring proposals denominated in cash: months to payback, contribution margin per visit, clinician hours released, support tickets removed. The three shipped AI features are, tellingly, all cost-side: charting time, chart-review time, receptionist labour [JD] [PRESS-2026]. That is the shape of a company that funds itself. Any growth-side bet has to be argued in the same currency, and the fastest credible ones are in the seams this Area exposed — a second live consumer brand with a different state footprint [BRAND], and an entire "AI Clinic Track" whose commercial model is undefined in public.

The third fact is the geography, and it is the operating risk nobody names in the posting. Every non-clinical requisition for the last nine months has been in Poland, Serbia, Spain, Portugal, Armenia, Georgia, Estonia, Cyprus, Italy, Turkey, Pakistan and Latin America; not one US engineering, product, design or data role appears anywhere in the board history I pulled [ATS] [ARC]. Screening asks whether the candidate is comfortable with CET or GMT+4 [ATS]. The patients, the prescribers, the pharmacies, the DEA and the state boards are all American. That is a permanent 6-to-11-hour gap between the people who build the product and everyone the product acts upon — and a Glassdoor review already reports "language barriers among support staff who have difficulty understanding provider needs" [REV]. A new VP of Product should treat this not as a cultural grumble but as a design constraint: the org cannot learn from US clinicians and patients synchronously, so it must learn from them instrumentally. Which is a polite way of saying the data platform is load-bearing — and there is exactly one data engineer for six analysts and a director [JD-SITE]. That is the first hire to argue for, and the first thing to say out loud in the interview.

The fourth fact is that the org's shape tells you what it is bad at. Eight PMs over 30 engineers is 3.75 engineers per PM [REASON] — an organisation configured to investigate rather than to ship. One analyst per PM and one data engineer per seven analysts [REASON] means answers are produced by hand and probably do not agree with each other. Three designers across two tracks and two consumer brands is starvation, which is why a Head of Product Design was hired in July 2026 [ARC]. And a full director layer exists above capacity that does not yet exist beneath it. The honest first-90-days read is: this company does not need more analysis, it needs fewer, larger bets with instrumented outcomes — which means consolidating squads, defining a small number of metrics that everyone computes the same way, and resisting the temptation to give each of eight PMs a domain. Expect resistance, because the existing director layer's authority is currently expressed through breadth.

The fifth fact is the honest read on the offer, and it has to be handled directly. The company's own posting is a Contract, B2B, Poland/Remote-EU engagement with a probation period and Polish-market benefits, and it does not mention equity; the recruiter-authored copy Ilia was sent promises "competitive compensation and equity aligned with impact" [JD] [JD-SITE]. Benchmarked against the Polish director/C-level B2B market, the plausible cash band is roughly US$106k–233k net-of-nothing, versus a US equivalent customarily in the US$250–350k base range plus equity — a 1.5x–3x cash gap [COMP-PL] [REASON], all of it estimate. The trade being offered is explicit in the posting's own words: "a scope typically reserved for C-suite executives at much larger companies" [JD-SITE]. That trade can be worth taking — the mandate is genuinely broad, the AI surface is genuinely shipped, and the business genuinely funds product work from operations. But it is a trade, and the equity discrepancy is the one thing that must be resolved before any of the rest matters, because in a founder-controlled company with no institutional lead and no exit clock [OWN], contractor equity is a low-probability instrument even when it exists.

What a new VP of Product should leave alone. Clinical protocol content — that is the P.C.'s and the medical director's, and touching it converts a product problem into a regulatory one. The MSO/P.C. structure itself — it exists because the law requires it, and re-litigating it wastes the only political capital worth spending. The clinician-recruiting machine — 35 of 59 requisitions and 33 of 35 states [ATS] [REASON] say MedOps has this running as a permanent funnel, and product's contribution there is credentialing and scheduling throughput, not recruiting strategy. And, at least for two quarters, the second consumer brand: EZCare is a live revenue surface with its own analytics [BRAND], and consolidating brands is a six-quarter programme that will eat everything else. Note it, price it, do not start it.

The one tradeoff to name out loud in the room. This company's entire product thesis is automating the routine follow-up — the CEO says up to 80% of psychiatric visits are routine follow-ups and describes a future of "fully automated follow-up care for stable patients" with "physicians in a supervisory role" [PRESS-2026]. The follow-up is also the revenue: it is the $159 recurring visit [WEB]. So the more successfully product automates the visit, the more directly it attacks the unit of sale. Every AI initiative is simultaneously a margin lever and a revenue-model question, and nobody can hold both sides of that except the person who owns product and the data and the funnel — which is precisely the job being advertised. A candidate who names that tension, and offers a view on whether the follow-up becomes cheaper, becomes a subscription, or becomes a supervised-automation product with a different price, will be the only person in the room arguing about the actual business.

The five questions only someone who did this work would ask.

1. "Your careers page lists this role as a B2B contract in Poland or Remote-EU, with 22 vacation days and no mention of equity; the brief I was sent says 'competitive compensation and equity aligned with impact.' Which document is the offer — and if there is equity, which entity issues it, on what instrument, at what last priced valuation, and what happens to it if the contract ends?" This is the honesty test and the economics test in one, and both documents are public, so it cannot be deflected.
2. "Alex Kulitski is described as MEDvidi's CTO on Crunchbase, on his own LinkedIn, and in your 2021 joint press release with Smart IT — but never on medvidi.com, and the posting puts a VP of Engineering under me. Today, who decides engineering direction, and does the VP of Engineering report to me or to Alex?" This is the question that determines whether the job is real.
3. "Smart IT's own case study says it was 'responsible for the entire development of MEDvidi' and supplied a CTO as a service, and your recruiters use smart-it.io addresses. Of the 30+ engineers I'd inherit, how many are MEDvidi contractors and how many are Smart IT's, and which entity signs their contracts?" This determines whether he can actually re-shape the team.
4. "You published a Director of Revenue Cycle Management role in July, your member terms authorise you to bill insurers on a patient's behalf, and Smart IT listed 'insurance claim management' as in progress years ago — while every consumer page says you accept no insurance. Is a payer-billed line already running, and if not, is it in the two-quarter plan?" Nobody who read only the marketing site can ask this, and the answer reframes the whole product agenda.
5. "Six analysts, one data engineer, and eight PMs — plus a requirement that I diagnose by segment, root cause, fix. How do you currently agree on the definition of a follow-up, an active patient, and a churned patient across brands, and who owns that definition? And relatedly: EZCare lists twenty states including New Jersey and Oklahoma, which MEDvidi doesn't — how is state eligibility modelled across the two brands?" This is the question that reveals whether the data layer can carry the AI strategy, and it lands squarely on the "segment → root cause → fix" requirement the posting itself set.