— Proof & Definitions

How we define our metrics.

We hold to one rule across this entire site: every claim is either defined here or labeled to its source. No performance figure without its population, period, method, and limitations; no client logos implying outcomes we won't substantiate. This page exists so you can check our numbers against their definitions.

— The standing rule

Career figures are career figures.

The headline numbers you may see associated with GenovateAI — 7 exits, 30+ operators, 25+ years, 100+ companies — reflect Liron Langer's full career as an operator and investor, including his role as CIO at the Nielsen Innovate Fund. They are not a GenovateAI fund record, and GenovateAI does not present them as one. GenovateAI is an advisory and execution firm; it does not operate a fund and publishes no fund-level returns.

Market figures we cite — $4.4T (McKinsey, 2023 — annual generative-AI potential) and 95% (MIT, The GenAI Divide, 2025) — are third-party estimates, labeled to their sources wherever they appear. We publish internally measured aggregate performance figures only when the population, denominator, reporting period, definitions, exclusions, calculation method, evidence owner, and material limitations are disclosed; such figures are identified as internally measured unless independently attested. Sample artifacts on this site are explicitly marked illustrative/redacted, and we do not represent client outcomes.

— Definitions

What each term means

Exits (7)

Counts companies that reached a liquidity event — acquisition or other realized exit — across Liron Langer's full career as an operator and investor, including his tenure as CIO at the Nielsen Innovate Fund. This is a career figure, not a GenovateAI fund track record. GenovateAI does not operate a fund and publishes no fund performance.

Operator network (30+)

Senior operators — former CTOs, CPOs, CIOs, GMs — Liron has worked with directly across portfolio companies, accelerators (8200 EISP, Techstars), and prior roles, who can be drawn on for an engagement. “Network” means people with a working relationship, not a mailing list or LinkedIn count.

25+ years

Liron's cumulative years operating, building, and investing in technology — a full-career figure spanning operating roles, fund leadership (incl. CIO at Nielsen Innovate Fund), and advisory. It is not the age of GenovateAI.

100+ companies

Companies Liron has evaluated, advised, invested in, or operated alongside over his career, again including his fund and accelerator roles. A career figure — not a count of GenovateAI clients or GenovateAI outcomes.

$4.4T — generative-AI value (McKinsey)

A third-party market-sizing estimate published by McKinsey for the potential annual economic value of generative AI. McKinsey's underlying report estimates $2.6T–$4.4T of potential annual value across the 63 use cases it analyzed — $4.4T is the upper bound of that range. It is labeled as McKinsey's figure wherever it appears and is not a GenovateAI projection or result. Source: McKinsey, 2023 →

95% — Enterprise GenAI pilots with no measurable P&L impact (MIT Project NANDA, 2025)

A third-party figure from MIT's The GenAI Divide: State of AI in Business 2025 (Project NANDA), which found ~95% of enterprise generative-AI pilots deliver no measurable P&L impact. It is labeled to its source and is not a GenovateAI-generated or client-specific number. Source: MIT Project NANDA, 2025 →

— Evidence levels

How we classify evidence

1 · Stated

A claim we make without published support. We aim to have none of these on the site; if you find one, treat it as an error.

2 · Documented

Defined and sourced — career figures, third-party statistics, and methodology pages like this one.

3 · Demonstrated

Shown working — sample artifacts (illustrative, redacted) and composite case studies drawn from real engagements with details anonymized and composited.

4 · Measured

A defined metric with baseline, period, sample, and limitations published alongside the result. We publish measured claims only with their methodology.

5 · Attested

A named client or independent party has approved the statement. We arrange reference calls during Discovery rather than publishing anonymous endorsements as attested proof.

Where this site stands today: samples in Tools & Samples are level 3 (illustrative); composite case studies are level 3 (demonstrated, composite); statistics and career figures are level 2 (documented). Levels 4–5 are published only as they are earned — the first level-4 instance (a measured internal workflow, anonymized, n=1) is published on Meta3Agents.

— Why this page exists

You should be able to trust every number on a page that asks you to trust a firm with your AI decisions. If a figure here isn't defined or sourced to a named third party, treat it as an error and tell us — we'll fix it.

Sample deliverables across Tools & Samples are illustrative and redacted; they do not depict real client engagements, names, or results.

Book an AI Decision Call

Questions about how we define a metric, or want the basis behind a figure? Raise it on the call — we'd rather over-explain a number than overstate it.

Book an AI Decision Call