Data Scientist
First seen by Alion on Oct 27, 2026.
About Lightly
Lightly is a deep-tech startup building an AI forensic intelligence platform. From individual chemical samples, it pieces together connections between cases and helps tell whether seemingly unrelated cases are in fact linked. Our data comes from Catcher, our own portable device built on patented UV-Fingerprint technology, which creates a unique profile of an unknown sample within minutes, right in the field. Every measurement feeds a shared database, so we don’t just say what’s in a sample, we can tell where we’ve seen something like it before and what it might connect to. We work with law-enforcement agencies in the Czech Republic, the EU and the U.S.
About the role
You will own the path from a measured sample to an answer you can trust: the evaluation algorithms, the quality of the data behind them, and the design of the measurements themselves. Most of the problems you’ll face haven’t been solved before, so solid mathematics and statistics are a must. And we’re a startup: one week you’re tuning a model, the next you’re planning a customer pilot or measuring samples in the lab yourself.
What you’ll do
Develop and benchmark methods for evaluating spectral data
Keep our measurement database clean and our experiments reproducible
Design measurement campaigns and pilots, evaluate them and report clearly
Measure substances yourself from time to time, both in the lab and at partner sites
Who we’re looking for
A strong foundation in mathematics or physics
At least 2 years of experience working with data, machine learning or research (a PhD counts)
Strong Python and practical SQL
Experimental design and honest validation
Willingness to travel in Czechia and abroad; fluent English
Nice to have
Signal processing, spectroscopy or chemometrics, lab experience, driving licence (cat. B).
What we offer
Real data from our own instruments, not public datasets
Work with real impact: what you come up with decides what the device recognises in the field
The whole chain, from measurement to deployment, in one role
A small team, short decision paths and work with law-enforcement agencies
How to apply
Send your CV and a few sentences about a project with real (and messy) data.