Research Engineer

Research Engineer

New York City  ·  On-site  ·  Full-time

Reports to: Sr. Research Scientist & CPTO

ABOUT ASTERA

Astera is building an AI-powered consumer platform for events. We study how real-world events evolve and propagate through markets, and build causal and world-model AI systems that reason and act on that structure in real time — shipping research directly into a live product used by real people.

ABOUT THE ROLE

We are looking for a research engineer to build the ML infrastructure that powers Astera’s AI systems. You will sit at the intersection of research and engineering — implementing novel model architectures, building training and evaluation pipelines, and ensuring that cutting-edge research runs reliably at scale in a live product.

This is not a role where you maintain existing systems. You will build new ones. You'll work side-by-side with research scientists, translating ideas into code and code into production. The right person is a strong engineer who is deeply curious about ML and wants to be close to the research frontier without needing to define the research agenda themselves.

WHAT YOU WILL DO

• Implement and optimize model architectures designed by the research team — world models, causal reasoning systems, multimodal encoders
• Build and maintain training, evaluation, and inference pipelines for ML models in production
• Design data pipelines for real-time, event-driven financial data across multiple modalities (text, time series, structured data)
• Optimize model performance: latency, throughput, memory, and cost
• Build internal tools for experiment tracking, model comparison, and debugging
• Collaborate with research scientists on rapid prototyping and with product engineers on production integration

YOU MAY BE A GOOD FIT IF YOU

• Have 1+ years of experience in ML engineering, research engineering, or a strong ML-focused academic background
• Are proficient in Python and PyTorch, and can read and implement ML papers
• Have experience building data pipelines, training loops, and evaluation harnesses
• Care about code quality — your systems need to be correct, fast, and maintainable
• Are curious about the research side of ML and want to work closely with researchers
• Thrive in fast-paced, ambiguous environments where you own problems end-to-end

STRONG CANDIDATES MAY ALSO

• Hold a BS or MS in CS, ML, applied math, or a related field
  Have contributed to ML research projects (co-authored papers, research internships, or open-source ML tools)
• Have experience with distributed training, GPU optimization, or model serving at scale
• Have worked with financial data, time series, or real-time streaming systems

HOW WE WORK

We are a small, research-driven team where everyone has outsized impact. Engineers do research. Researchers ship code. The distance between an idea and production is measured in days, not quarters. We value clarity of thought, speed of iteration, and strong research taste — the instinct for which problems matter and which experiments will actually teach you something. We are not interested in incremental work. We want people who think from first principles, move fast, and care deeply about craft.

HOW TO APPLY

Send your CV, a link to your best technical work (GitHub, papers, or projects), and a brief note on why Astera to akshay@astera.holdings.