Data Scientist · Buffalo, NYOpen to data science roles at startups

I find one small problem at your company and prototype the fix.

Each project here takes one small problem at a real company and builds a working prototype of the answer, on sample data. Open any of them, and tell me if I got it wrong.

Before this, in production

8.3%revenue forecast error, one pooled model replacing nine, adopted for budget planningBluelight (YC)
$14M+in invoices turned into technician benchmarking behind raises and bonusesBluelight (YC)
10,000+athletes using the AI Coach feature I built end to endBrandlete
150Kleads scored to show which marketing campaign won whereCoverRight

Case files · 11

One company, one prototype.

Each one is a clickable prototype on sample data, built before anyone asked. Start with these three.

All other case files, by what they solve8

Method

How this works.

01

I pick one problem

Something small and real at a company I'd like to work with.

02

I build a prototype

An interactive concept, not a deck: sample data, method and result you can click through.

03

You open it

If it's useful, we talk. If I got it wrong, I'd like to know.

Anirudh Devella
Anirudh Devella · Buffalo, NY

About

I build forecasting, scoring and benchmarking systems, end to end, from warehouse query to delivery.

Data scientist with about two and a half years across product and go-to-market analytics, after three and a half years as a software engineer. I build forecasting, scoring and benchmarking systems in Python and SQL, end to end from warehouse query to delivery. M.S. Data Science, University at Buffalo.

The full story →

Contact

Have a problem in mind?

Tell me what's slow, leaky or guessed at, and I'll build a first prototype.

anirudhdevella@gmail.com