Ari Leshno

Role

Ophthalmologist and glaucoma specialist at sheba; also head of the eye institute’s research-and-innovation unit. Has a Sheba-backed budget + green light to fund a healthcare startup and is actively recruiting entrepreneurs. Met saar-arbel and guy-barkat on 2026-05-19 to pitch a glaucoma visual-field tracking wedge — see 2026-05-19-saar-leshno-glaucoma-data-tracking.

Affiliations

• sheba — clinical (glaucoma specialist) + administrative (head of research and innovation, eye institute). Distinguishes between “how I work in Sheba” and “how I work in private practice”; also runs a private clinic on the side.

Notes

• Introduced to saar-arbel by “Ohad” (no surname captured); Ari is ~13 years younger than Ohad and resembles him. Ohad is the connector who flagged Ari to Saar.
• Domain expertise: 13 sub-fields within ophthalmology; glaucoma is his specialty. Fluent on both the technical (Humphrey perimetry, OCT, retinal imaging) and political (Zeiss instrument lock-in, FDA gating) layers.
• Concrete pitch (2026-05-19): OCR-based cross-clinic ingest of visual-field test PDFs + a better tracking UI; subscription model ~$100/clinic/month;sizinghegave:35Kophthalmologyclinicsglobally\times 10$100/mo ≈ tens of millions/year. His resident yael built the MVP in Claude in ~30 minutes. Sheba holds a multi-year raw dataset (Excel exports + scanned PDFs from years of glaucoma practice) and would route OCR development through Sheba’s internal AI center.
• Two explicit warnings to Saar/Guy: (1) avoid the AI scribe / note-taker space — saturated; (2) avoid diagnostic AI as a starting point — every diagnostic claim is an FDA event. Reference for the workaround: idx / “AI-Doc” got around FDA by phrasing output as “this scan is normal” rather than diagnosing.
• Future-state plays (FDA-gated): deterioration prediction, treatment recommendation, missed-stroke detection on retinal imaging. Sheba is already developing comparable tooling internally — both a competitive concern and a validation signal.
• Used in 2026-05-27-directions-vertical-pivot-and-prediction-markets (one week after this pitch) as the anchor anecdote for vertical-use-case-led-brain — Saar’s wording: “we already started thinking about how to bend the [Brain/MCP] tool to fit his problems.”
• Personality: opens with an explicit “I’m bad with names and faces — apologize in advance” disclaimer.

Mentioned in

• 2026-05-19-saar-leshno-glaucoma-data-tracking — pitched the visual-field tracking wedge; offered Sheba budget + entrepreneur partnership; gave the don’t-touch list.
• 2026-05-27-directions-vertical-pivot-and-prediction-markets — anchor anecdote for the vertical-Brain pivot argument (cited as hari in the 2026-05-27 ingest, before the rename refactor merged that page into this one).