Lili.ai's bet is that the risks and disputes of large construction and energy projects can be spotted early in the projects' own documents. Lili.ai, a Paris company founded in 2016 by Milie Taing, uses artificial intelligence to read this unstructured material, flag risks early and gather evidence for claims.
The European Innovation Council granted €2,591,604 towards a €3.70 million project that, according to CORDIS, delivered a production-ready platform. In 2021 Lili.ai said €3.4 million of matching equity was planned for 2023. No round, EIC Fund investment or customer has been announced since.
Company file
- Legal entity
- Lili.ai, SIREN 820982627, Paris, France
- Founded
- June 2016, by Milie Taing (CEO)
- Team
- CTO Gilles VanWormhoudt; chief scientist Nadi Tomeh
- Product
- AI platform for risk monitoring, claims and document analysis
- Markets
- Construction, energy, transport, defence
- Ecosystem
- Oracle Construction and Engineering Inno Lab; integrates with Oracle Unifier
- EIC project
- Lili.AI, 1 Jun 2022 to 31 May 2025
- EIC funding
- Blended finance
Where Lili.ai Stands Today
Lili.ai's recent posts describe events rather than business results. In April 2025 Milie Taing spoke about the company's EIC experience at a French government funding day, arguing for AI "grounded in Europe's unique assets". In October 2024 it showed how its platform integrates with Oracle Unifier at an Oracle partner summit. The latest accounts in the French register data, for 2022, record revenue of zero and a small profit of €23,245.
The Technology: Reading the Project File
According to its website, Lili.ai processes unstructured data such as documents, emails and PDFs that VBA macros cannot handle. The CORDIS report describes a big-data platform with machine learning and language models for classifying documents and extracting entities, semantic search and risk monitoring, deployable on several clouds, and calls it "a robust, multilingual, real-time AI platform".
"Equip all major projects in their ongoing transition from litigation to early identification and resolution of disputes."
Lili.ai, on its EIC award, December 2021What the EIC Project Promised, and What CORDIS Shows
The project promised a "disruptive AI project management solution to drastically reduce the costs & risks of European large-scale construction projects". CORDIS reports the platform reached technology readiness level 9, meaning proven in operation, and cites an average construction claim of about €54 million as the cost it aims to reduce. It reports no clients, deployments, case studies or revenue.
- A production-ready platformDeliveredTRL 9, per CORDIS.
- Integration with project toolsDeliveredOracle Unifier demo, 2024.
- Customers on large projectsNot shownNone named.
- Matching equityNot shownPlanned for 2023; not announced.
What Has Happened Since the Award
The Money
The award was blended finance. Lili.ai said in December 2021 that €3.4 million of matching equity was planned for 2023. We found no announcement of that round, no EIC Fund portfolio listing and no named investors. Its 2022 accounts show no revenue.
What cannot be checked
Who uses the platform, whether the EIC took equity, Lili.ai's revenue since 2022 and the size of its team are not public.
Why Europe Should Care
Lili.ai presents itself as European industrial AI built on European data. The EIC funded that pitch and got a finished platform. Whether large contractors and project owners pay for it is not visible in any public source.
What the project promised
AI to drastically reduce the costs and risks of Europe's large construction projects.
What the record shows
A TRL 9 platform and an Oracle integration; no named customers, no revenue in 2022 and no equity round.
The Verdict
Built, but unseen in the market
The EIC grant produced the platform it paid for. What is missing is any public evidence that it is used on the large projects it was built for, and the equity round meant to finance growth never appeared. Lili.ai should publish its customers if it has them.
Lili.ai has not been asked for comment for this article. It is based entirely on company statements, EU project data and the French company register, linked throughout.
