Skip to content
COLD START / catalog / c/datologyairetrieved 2026-06-23T16:30 UTC · 15 sources read

DatologyAI

DatologyAI automates data curation for AI model training, helping teams train better models with less compute.

datologyai.com · Redwood City

CALL NO.CS·DATOLOGYAI·26
StageSeries AMay 2024 · led by Felicis Ventures[1][11][3]
Raised$58Mdisclosed rounds[12][11][1]
Headcount43as of Apr 2026[12]
Valuationnot publicly disclosedno source in ledger
Founded2023[12][7]
Money

Raised $58M across disclosed rounds.[12][11][1]

Series A closed May 2024, led by Felicis Ventures.[1][11][3]

post-money valuation is not publicly disclosed.

Named investors include Felicis Ventures, Amplify Partners, Radical Ventures, M12 (Microsoft), Amazon Alexa Fund, Elad Gil, Conviction Capital, Outset Capital, Quiet Capital.[1][2][4][11]

People

Ari Morcos, CEO and Co-Founder[8][9][10]

Matthew Leavitt, Chief Science Officer and Co-Founder[8][9][10]

Bogdan Gaza, CTO and Co-Founder[8][9][10]

Signals
Jun 23 2026newsDatologyAI (United States) company claim, not independently confirmed[12]
Jun 23 2026newsAI Training Data Curation Platform | DatologyAI company claim, not independently confirmed[13]
Jun 23 2026newsThomson Reuters Improves Model with DatologyAI’s Data Curation company claim, not independently confirmed[14]
Jun 23 2026newsResearch Scientist, Post-Training at DatologyAI | Redwood City company claim, not independently confirmed[15]
May 20 2026launchDatologyAI publishes VLM results showing 11.7pp improvement through data curation alone[8]
Apr 9 2026launchThomson Reuters partners with DatologyAI for legal domain model training[8]
Comps

CleanLab · cleanlab.ai · Same buyer (AI teams) and same workflow (data curation for model training), but CleanLab focuses on label quality and data error detection rather than petabyte-scale pretraining data selection.[2]

Labelbox · labelbox.com · Same buyer (AI teams) and adjacent workflow (data preparation), but Labelbox is primarily a human-in-the-loop labeling platform, not automated petabyte-scale curation.[2]

Galileo · rungalileo.io · Same buyer (AI teams) but different workflow: Galileo focuses on model evaluation and observability rather than pretraining data curation.[2]

YData · ydata.ai · Same buyer (AI teams) and adjacent workflow (data quality), but YData focuses on synthetic data generation rather than selecting from existing petabyte-scale datasets.[2]

Risk

The customer proof is company-sourced only. No independent source in this ledger confirms it.[14]

Next question

Ask for one referenceable production customer.

Not a recommendation. The first thing this ledger cannot answer.

Investor Lens
Why care
The case
What to learn next
Pay attention to
How it wins

Filled in the side panel for invited readers. This card carries sourced facts only.

Sourcestracings · 15

DatologyAI

CALL NO.CS·DATOLOGYAI·26

DatologyAI automates data curation for AI model training, helping teams train better models with less compute.

datologyai.com · Redwood City

StageSeries A
Raised$58M
Headcount43
Valuationnot publicly disclosed
Founded2023
CompsCleanLab +3

Ask for one referenceable production customer.

Not a recommendation. The first thing this ledger cannot answer.