We're looking for an entrepreneurial data scientist/quant

Hey, we're Sebastian & Rayane, the founders of Kairos.

We're building systems that predict sports outcomes (accurately, we hope). To do this, we're putting together a core team of highly technical quants/software engineers/quant devs, one of which could be you!

What we actually do (and why it's fun)

Let's say Mbappé is slightly injured but he might still play tonight's game. What's the impact of the injury report on the probability that he scores? What's the impact on the overall probability of Real Madrid winning the game? Or let's say Djokovic is going to play the US Open semi-final against Alcaraz at 3pm local time. It's probably going to be quite warm at around 30°C. How does that impact his chances of moving onto the final and going for a 25th Grand Slam?

We try and give good answers to these kinds of questions. We ingest lots of data, build models to make sense of all of it, and generate forecasts. We then use these forecasts in systematic and quantitative trading strategies.

Your role: data scientist/quant researcher to make our models less wrong

What you'll work on

• Building and improving forecasting models for sports outcomes, pre-match and in-play, across sports and market types
• Turning these forecasts into quantitative trading strategies: backtests, sizing, risk management
• Owning models once they're live: monitoring calibration, drift, edge decay
• Working with a lot of real, messy data and thinking about production with the eng team

What might be useful

Research here doesn't stop at a Jupyter notebook. Quants think about the production design and work collaboratively with the engineering team. Here are some things that might be useful (we don't expect you to master each area):

• 4+ years building models that went into production, or a PhD plus real applied work. Strong Python knowledge and experience with DS libs (numpy, pandas/polars)
• A quantitative background - stats, ML, applied maths, physics - you should be able to argue about a forecast
• Hands-on experience with: probabilistic modeling, bayesian and hierarchical models, simulation engines
• Experience with machine learning models and neural nets (e.g. LightGBM/XGBoost, PyTorch), along with knowing when a simple model wins
• Healthy paranoia about evals, overfitting, and leakage + the ability to write backtests you trust
• Familiarity with market data: orderbooks, fees, slippage and execution on exchanges
• SQL and being able to pull your own data from a lakehouse (Parquet/Iceberg, time-series DBs)

Are we a good fit?

We're an early-stage startup, working in-person from Paris. So beyond the kind of obvious traits that mean you'd enjoy working on the above you need to be pretty comfortable with uncertainty. So, you should join us if:

• Early-stage ownership sounds exciting, you're comfortable with taking on more risk for a big upside
• You value intellectual freedom and personal agency
• You want to work with smart people, who move fast
• You're competitive (in sports, games, …)
• You want to grow with the company - both professionally and financially (we offer competitive pay and real ownership as this will be your baby too)
• Bonus: you enjoy sports, company events may include sports games & we often practice sports together

How to apply

Email us at draftpicks@kairos-research.ai with your CV / portfolio. Even better if you have an anecdote around sports and/or forecasting 😁

Don't self-reject! You don't need to know everything about sports, domain knowledge is easy to teach.

PS If this doesn't sound like you but you know just the right gal/guy, we'd appreciate the intro (and buy you 2 tickets to your favorite sports event)