François Chollet

Researchers / independent engineers · Software/Tech · milestone at age 25 ·T2 Field-leading

Milestone (age 25)

Created and released Keras, a widely adopted high-level deep learning library for Python, around February–March 2015 at age 25; joined Google shortly after.

French engineer born 20 October 1989; earned a Diplôme d'Ingénieur (MEng) from ENSTA Paris (Polytechnic Institute of Paris) in 2012. Built Keras as a side project in early 2015 during the pre-framework-standardization wave of deep learning, then joined Google where Keras became tightly integrated with TensorFlow.

Think your path resembles François Chollet's?Compare the visible ingredients, then see exactly where the comparison stops working.

Am I the next François Chollet? →

Starting point

Born 20 October 1989 in France; trained as an engineer at ENSTA Paris (MEng 2012).

Current position (2026)

AI researcher and entrepreneur after leaving Google (2024); co-founded AGI-focused lab with Zapier co-founder Mike Knoop; creator of Keras and ARC-AGI / ARC Prize.

How this path compounded

01 Starting advantages

6/24 starting-position score

Strongest documented signals: Elite institution pipeline, Early online platform, Frontier geography.

Describes the starting position, not what the person later made of it.

Cohort percentile: 41

02 Built or converted leverage

13/25 multiplying-capacity score

Strongest observed levers:Domain proximity, Prior reps, Scarce skill depth.

Measures what was present, not whether it was inherited, earned, self-built, external, or mixed.

Cohort percentile: 55

03 Compounding trajectory

6 documented steps

The timeline below shows the sequence of work and transitions around the selected early milestone. It is evidence of a path, not proof that every step was necessary.

Milestone at age 25

04 Observed career standing

T2 · Field-leading

Dominant figure at the top of a field. The tier summarizes documented career recognition through the data cutoff—not François Chollet's worth or future potential.

How tiers work →

Question four · where did the leverage come from?

François Chollet's leverage provenance

Each non-zero lever gets a best-supported origin, evidence signals, and confidence. Unresolved is the honest default when the biography cannot distinguish self-built, advantage-enabled, earned, external, or mixed.

Domain proximity2/2

Advantage-enabledmedium confidence

One or more documented starting advantages plausibly enabled this lever.

Direct domain exposure (1/2)Frontier geography (1/2)Elite institution pipeline (2/2)

Prior reps2/3

Advantage-enabledmedium confidence

One or more documented starting advantages plausibly enabled this lever.

Elite institution pipeline (2/2)Early online platform (2/2)

Scarce skill depth2/3

Advantage-enabledmedium confidence

One or more documented starting advantages plausibly enabled this lever.

Elite institution pipeline (2/2)Early online platform (2/2)

Elite ecosystem network2/3

Advantage-enabledmedium confidence

One or more documented starting advantages plausibly enabled this lever.

Elite institution pipeline (2/2)Frontier geography (1/2)

Structural wave / timing2/3

Externalmedium confidence

A structural wave is external to the person, even when their position improved access to it.

Frontier geography (1/2)Early online platform (2/2)

Concentration intensity2/3

Unresolvedlow confidence

No current annotation distinguishes self-built, enabled, or earned origins for this lever.

No decisive linked signal

Native distribution1/3

Advantage-enabledmedium confidence

One or more documented starting advantages plausibly enabled this lever.

Early online platform (2/2)Elite institution pipeline (2/2)

This is a bounded inference from the current annotations—not a claim about private effort, merit, or the percentage of François Chollet's outcome attributable to any origin.

Luck is not a leftover score.

Structural luck, Encounter luck, Event luck, Outcome variance can change every arrow in the path. This successful-only dataset cannot observe the near-identical paths that did not break through, so luck stays visible and unscored.

Within Researchers / independent engineers, François Chollet's starting-advantage total is at the 41th percentile. Separately, their built or converted leverage total is at the 55th percentile. Other T2 profiles average 7.9 / 24 starting advantage and 12.3 / 25 leverage. Similar scores appear in other tiers, so these figures describe positioning—not a cause.

Trajectory

2012 · age 22
Graduated ENSTA Paris with a Diplôme d'Ingénieur (MEng).
2015 · age 25
Released Keras deep learning library and joined Google shortly afterward.
2017 · age 27
Published Xception paper and Deep Learning with Python book
Amplifying Keras adoption.
2019 · age 29
Published ARC-AGI benchmark paper On the Measure of Intelligence.
2024 · age 34
Named to TIME100 AI; left Google
9+ years; launched ARC Prize competition.
2025 · age 35
Co-founded new AGI/program-synthesis lab and expanded ARC Prize into a foundation.

Primary leverage engine

Scarce technical / intellectual depth

Secondary engine

Product / domain insight

Built/converted leverage

13 / 25

evidence: High

Built or converted leverage

Multiplying capacity documented later in the path. Measures what was present, not whether it was inherited, earned, self-built, external, or mixed.

Started serious reps before 20

0/1

Prior reps

2/3

Scarce skill depth

2/3

Native distribution

1/3

Elite ecosystem network

2/3

Complementary team

0/2

Structural wave / timing

2/3

Concentration intensity

2/3

Capital safety

0/2

Domain proximity

2/2

Starting-advantage scores (0–2 each)

Access or conditions documented near the beginning of the path. Zero means "no clear evidence in reviewed sources," not "advantage was absent."

Family financial platform

0/2

Parent / family domain

0/2

Inherited audience / network

0/2

Elite institution pipeline

2/2

Frontier geography

1/2

Rare early tools

0/2

Dedicated mentor / coach

0/2

Exceptional peer / cofounder

0/2

Early online platform

2/2

Direct domain exposure

1/2

Prodigy / innate ability

0/2

Adversity / constraint catalyst

0/2

Family context

Not documented in reviewed sources.

Parent / family domain

Not documented in reviewed sources.

Archetype & tags

Institutional ecosystem accelerationENSTA Parisdeep learning wave 2015open-source KerasGoogle integration

Evidence summary

Chollet released Keras at 25 in the 2015 deep-learning boom, then joined Google where it became a default high-level API path into TensorFlow. Elite French engineering education plus open-source distribution and structural ML wave are the clearest advantages; family background is not documented in reviewed sources.

advantage confidence: High · source count: 3 · audit: not_independently_audited · status: subagent_researched_beta

Sources

Related — same primary engine

Mark Zuckerberg

Founder/Entrepreneur · age 19

Launched Facebook at Harvard

Leverage 19/25T1

Michael Truell

Founder/Entrepreneur · age 23

Co-founded Anysphere/Cursor

Leverage 19/25T1

Alfred Wahlforss