Machine Learning Engineer

Percepta • New York, NY

91d

About The Position

We’re hiring Machine Learning Engineers who will work directly within customer teams to define and deliver high-impact AI systems. We don't build prototypes or laboratory projects - you’ll design, build, and ship production-grade AI agents and workflows that drive millions in business value for customers.

Requirements

• AI-nativeness: You're excited about the potential for AI to transform businesses and want to play a hands-on role in bringing frontier technology into critical institutions.
• Strong ML foundations with hands-on experience building and deploying production models / AI systems.
• Being generative and collaborative: You love constantly jamming on new 'what if' ideas with teammates and partners to bridge applied engineering, product, and research efforts.
• Extreme ownership: You’re willing to jump in and love being the one on the hook.
• Execution excellence and speed: You can build stuff in messy environments and know how to get code written and shipped quickly.
• Customer-obsession and respect: You’re motivated by understanding customer pain points and iterating directly with end users to deliver wins quickly.

Nice To Haves

• Hands-on experience with LLM tooling (e.g., LangGraph, Mastra, Agents SDK).
• Experience fine-tuning, distilling, and deploying LLMs or other foundation models in production.
• Background in retrieval, RAG pipelines, or multi-step agent design (including tool use and human-in-the-loop systems).
• Strong engineering foundations in Python/TypeScript, cloud deployment (AWS/GCP/Azure), and modern MLOps/DevOps tooling.
• Prior startup or founding engineer experience, balancing craft, ownership, and speed.

Responsibilities

• Engineer and optimize AI/ML systems: Build end-to-end ML pipelines for data ingestion, training, evaluation, and deployment.
• Adapt and extend LLM models with fine-tuning, distillation, retrieval systems, and tool-use to solve domain-specific problems.
• Evaluate AI systems rigorously: Develop custom evaluations to ensure models succeed in real-world environments.
• Bring frontier methods into practice: Track the latest techniques in areas like RAG, tool use, multi-step agent orchestration, fine-tuning methods, and evaluation frameworks - and apply them to specific customer challenges.
• Collaborate across product and research: Partner with research and product teams to turn frontier techniques into production-ready features and workflows.
• Advance our core product: Encode the lessons from our customer engagements in our Mosaic product, consistently contributing reusable ML components, infrastructure abstractions, and performance improvements.

What This Job Offers