Hi Inov · Venture Capital Internship · 2026
from first principles,

Venture
Capital

"Hands-on, from entrepreneurs to entrepreneurs."
Paris · Lyon · Munich — Seed to Series B

A business school student with science background turns investor-in-training — bridging peer-reviewed science and seed-stage conviction. Focused on AI infrastructure, ClimateTech, and the software enabling Europe's deep tech ecosystem.

B2B Software AI ClimateTech Cybersecurity Defence

Why are you interested in venture capital more than something else?
"Seed-stage investing is one question: is this science actually credible, and is this the right team to execute on it?"

I chose VC because I want to be the investor who can tell when a technical claim is real, and then help the right founders turn that into products that matter.

VC is the role that combines 3 dimensions at once: technical judgment, early-stage company building, and financial thinking. The most extraordinary part of early-stage investing is to be the person who says yes at the moment a genuinely important breakthrough needs a partner.

The vaccines, the gene therapies, the grid software that makes the energy transition survivable — these didn't happen because the science was obvious. They happened because someone with enough taste, enough conviction, and enough genuine care chose to back the founder behind them. That is what I want to spend my career doing. Hi Inov's focus — B2B software in AI, ClimateTech, Cybersecurity, and Compliance — is precisely the territory where I believe the next decade's most consequential bets will be made. I want to be in the room making them.

Formative experiences
Hello Tomorrow
Deep Tech Evaluation at Scale
Sourced over 20,000+ deep tech startups across 13 verticals and evaluated 100+ in the AI, Energy and Climate sectors — learning to distinguish genuine technical claims from hype before a single line of financial modelling.
VentureBlick
Health-tech Deal Experience
Facilitated early-stage healthtech deals — understanding how decisions made at seed compound through every subsequent funding round.

As an investor, what kind of bets would you make on the AI revolution?
"The real AI bets are not in the model layer — they're in the physical infrastructure the models depend on, and in the vertical software that owns both the data and the workflow it runs on."

My conviction is that the alpha sits in the energy and data infrastructure that makes AI physically possible, and in vertical software that has achieved genuine lock-in by owning the underlying data.

Three conviction areas

Energy Resilience

AI data centres need near-perfect power reliability. Yet both the US and Europe are underinvested in storage and grid capacity. In Europe, hundreds of billions are needed in grid upgrades, with a large share of distribution grids already older than 40 years. At the same time, microgrids and BESS software are growing fast. I would look for the software layer that operates this energy system: grid optimisation, demand forecasting, and control systems where the customer cannot churn without taking real operational risk.

Vertical AI with Proprietary Data

Open-source model trajectories suggest: profits shift to whoever controls proprietary data and the physical workflow it runs on. Bioptimus in Paris is a good example in biotech: open-source model first, then build a unique dataset with top pharma companies that no one else has. Similar logic applies in finance tools like Rowspace, which become core intelligence layers once they ingest years of internal data.


What is your best use of AI in your daily life?
"I built a system where AI handles information retrieval and pattern-matching — so my analysis stays reserved for insights requiring human synthesis."

The system replicates — at personal scale — what the best VC firms do institutionally: systematic market mapping, continuous signal intake, and structured knowledge compounding. In a role where pattern recognition across sectors is the job, this is a direct productivity multiplier.

Three-part system
01

Sourcing

Perplexity and Comet for fetching and identifying early-stage startups through the webpage of different fund and incubators' portfolio, and automatically creating a table for outreach sequences — compressing days of manual research into minutes.

02

Knowledge Integration

Claude Code processes newsletters, research papers, and podcasts into a structured personal knowledge base: 145+ learnings across 13 deep-tech verticals.

03

Pattern Recognition

The system surfaces cross-disciplinary connections automatically — preserving human judgment only for original synthesis and conviction-building.