Forge AI Architect — Autonomous Agent Platform

Norstella; Dover (DE); Remote; USD 180,000 - 250,000; Full time

Job summary

Norstella is seeking a Senior Forge AI Engineer to design and operate a greenfield, autonomous agentic platform. You will work across orchestration, multi-agent coordination, MCP server design, RAG pipelines, and safety primitives to ship reliable production systems.

You will tackle challenges in agent memory, trust, cost attribution, and compliance within a regulated life sciences environment, while building scalable, observable tooling that supports fast iteration and safe deployments.

Qualifications

• 6+ years of software engineering experience, with at least 2 years building and operating LLM-powered or agentic systems in production.
• Hands-on experience with MCP servers or designing tool-use APIs with explicit contracts and versioning.
• Proficiency in Python and/or TypeScript; able to own production agent code from design through operation.

Responsibilities

• Design and build agents across the Forge chain — orchestration, tool-use surfaces, MCP server integrations, prompt and RAG pipelines.
• Own multi-agent coordination architecture: inter-agent trust boundaries, message schemas, partial failure handling, recovery.
• Evolve the MCP server layer — tool schema design, capability contracts, versioning for safe extension.
• Design agent memory and state management: context windows, long-term memory, state persistence across multi-step workflows.
• Own agent quality end-to-end: eval harnesses, behavioral benchmarks, continuous evaluation pipelines.

Skills

LLM systems

Agentic frameworks

Python

TypeScript

Prompt engineering

Orchestration

RAG pipelines

Memory architectures

Vector databases

LLM evaluation

AI safety

Tools

LangGraph

CrewAI

Autogen

Semantic Kernel

MCP

Job description

Norstella is seeking a Senior Forge AI Engineer to design and operate a greenfield, autonomous agentic platform. You will work across orchestration, multi-agent coordination, MCP server design, RAG pipelines, and safety primitives to ship reliable production systems.

You will tackle challenges in agent memory, trust, cost attribution, and compliance within a regulated life sciences environment, while building scalable, observable tooling that supports fast iteration and safe deployments.