Production ML Research Engineer – OCR & Inference

Datalab; New York (NY); On-site; USD 250,000 - 350,000; Full time

Job summary

Datalab in New York City seeks a Research Engineer to own problems end to end across our models, inference service, and product. You’ll take models from training through benchmarking into our inference stack and collaborate to ship them to customers.

Join a small team shipping the state-of-the-art OCR models like Chandra and contributing to open source tools used by frontier AI labs and Fortune 500s. You’ll balance rapid experimentation with robust, scalable solutions for document extraction.

Qualifications

• Shipped models in production and understand end-to-end lifecycle.
• Experience turning research insights into products used by customers.
• Strong proficiency in PyTorch and Python with deep learning fundamentals.

Responsibilities

• Train task-specific models (OCR, layout, text recognition, extraction); explore architectures.
• Profile and accelerate model inference across hardware setups (H100s, CPUs).
• Integrate models into our API and product, balancing model work with shipping velocity.

Skills

DL model training

Fine-tuning

Model evaluation

PyTorch

Python

Deep learning fundamentals

Benchmarking

Performance profiling

Data engineering

Startup experience

Tools

PyTorch

Python

Job description

Datalab in New York City seeks a Research Engineer to own problems end to end across our models, inference service, and product. You’ll take models from training through benchmarking into our inference stack and collaborate to ship them to customers.

Join a small team shipping the state-of-the-art OCR models like Chandra and contributing to open source tools used by frontier AI labs and Fortune 500s. You’ll balance rapid experimentation with robust, scalable solutions for document extraction.