The R&D lab that takes your AI from "what if" to "ready to ship."

R&D lab for hire. Start with an idea, a dataset, or a prototype. We'll deliver it ready to ship, publish, or present—in weeks, not months.

Tell us about your data See the work

End-to-end AI/ML delivery: framing → data → training → deployment → monitoring.
LLM & generative AI expertise: RAG, fine-tuning, agents, tool use, prompt systems.
Rigorous evaluation for LLMs, search, and ranking (offline + online).
Reliability + safety: hallucination controls, guardrails, privacy/security, red-teaming.
Clean handoff: production code, tests, metrics, dashboards, runbooks.

Trusted by teams building ambitious products

What we do

R&D sprints

Rapid exploration, prototype baselines, and hard-nosed feasibility checks. See our work.

LLM evaluation & guardrails

Prompting, rubric design, and reliability testing before production. Learn about our LLM chatbot work.

Search & ranking research

Offline/online metrics, ranking experiments, and user-impact analysis. Read our search engine case study.

Messy data science

We clean messy data, find meaningful patterns, and test hypotheses. Explore our small datasets research.

Forecasting & anomaly detection

We build models that predict trends and alert you to unusual data points.

Research-to-engineering handoff

Clear documentation, evaluation reports, and deployment-ready guidance.

How we work

We run 4–8 week sprints. You get shipped code, research findings, or go-to-market demos. Not reports about what might work someday.

1–2 weeks

Discovery & data audit

Understand what you have and where you want to go. We audit data quality, frame the problem, and plan experiments.

Problem scoping & success metrics
Data quality assessment
Technical roadmap & experiment design

2–8 weeks

Experiments & evaluation

Build it. We implement baselines, test approaches, and run evaluations to prove what works.

Model development & prototyping
Custom evaluation frameworks
Performance benchmarking & safety testing

2–8 weeks

Production handoff

Deliver it ready to use. Production code, documentation, monitoring, and team training.

Production-ready implementation
Monitoring, metrics & dashboards
Documentation & team handoff

The Team

Sergey Feldman is a Principal Research Scientist at the Allen Institute for AI (Ai2) and Head of AI at Alongside. He leads work on production-scale AI systems, including core infrastructure for Semantic Scholar and Asta. His expertise spans LLMs, machine learning, natural language processing, and scientific document analysis.

Sergey founded Data Cowboys to bring research-grade ML to organizations solving hard problems. Sergey holds a PhD in machine learning from the University of Washington and has been building AI systems since 2007.

Ilya Barshai is a Principal Research Scientist at McGraw Hill, where he applies AI and machine learning to educational technology. He brings over a decade of hands-on data science experience.

Ilya joined Data Cowboys in 2016, bringing a practitioner's approach to ML consulting focused on building solutions that actually ship. Before moving into data science, he spent eight years in risk and failure analysis of electromechanical product designs.

What clients say

Sergey and Ilya grasped what we do the fastest of anyone I’ve explained our startup to. They are professional, timely, and so fast with their thinking and output.

— Jay Goyal, Co-Founder, Actively Learn & Alongside

Sergey has a wonderful knack for quickly scoping a problem, making efficient decisions on model choice and parameters, and clearly communicating throughout the process.

— Apu Mishra, Lead Data Scientist, RichRelevance

Sergey Feldman is brilliant, collaborative, candid, and lightning-fast. He has a special gift for explaining complex analytical concepts in terms anyone can understand.

— Deb Arnold, Principal, Deb Arnold, Inc.

Sergey provides well-considered results quickly with in-depth knowledge communicated clearly to readers at every level.

— Jenny Dearborn, Chief People Strategy Officer, BTS

Data Cowboys’ work has been incredible. We understand our data, verify our impact, and keep people safe from exploitation.

— Robert Beiser, Technology Operations Lead, Microsoft

From the blog

Which ML Classifiers Work Best for Small Datasets…

We tested 108 datasets to see what really works with 100–1,000 samples.

Underspecification in Machine Learning

Why models that perform identically in training can behave wildly in the wild.

Building Search for 180M+ Papers

How we improved Semantic Scholar search using three years of logs.

SHAP Values & Feature Variance

Making ML results understandable to collaborators.

Talks & Interviews

Building AI Products That Ship

Interview on moving ML systems from research to production.

The Data Scientist Show

Discussion on machine learning for search, consulting for Gates Foundation, AI for mental health, and career lessons.

Machine Learning for Global Health

Predicting child mortality with the CHAIN network using XGBoost and SHAP values to identify key physiological drivers for targeted interventions.

You Should Probably Be Doing Nested Cross-Validation

PyData Miami 2019 talk on double cross-validation for model selection and performance estimation on small datasets.

Got a data challenge…

Let’s figure out what’s possible before you invest in full buildout.

Prefer email… Reach us at ilya@data-cowboys.com.