Agent Infrastructure Engineer — Core Harness (Superagent)
ImagineArt · Remote · India ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Fully remote
- Python or TypeScript
- LLM Agent Infrastructure
- 4+ years experience
- GenAI Product
ImagineArt is a fast-growing GenAI company that has built a top-ranked image generation model without external funding. You will own Superagent, the infrastructure powering conversations and multi-step workflows across their platform.
This is a deep systems role involving the agent execution loop, memory management, and observability. You will optimize for latency, cost, and reliability while integrating multiple LLM providers.
What you'll do
- Own architecture and evolution of the Superagent harness
- Optimize agent loops for latency, cost, and reliability
- Build context management, memory, and tool routing systems
- Create evaluation infrastructure to measure agent quality
- Benchmark multiple LLM providers for performance and cost
- Implement caching, batching, and prompt compression
- Add deep observability including tracing and metrics
- Debug complex issues in non-deterministic distributed systems
What you'll need
- 4+ years in software, backend, or systems engineering
- Strong proficiency in Python and/or TypeScript
- Production experience building LLM-based agents
- Knowledge of tool calling, schemas, and structured outputs
- Experience with agent frameworks like LangGraph or CrewAI
- Understanding of agent orchestration and multi-step workflows
- Experience with evaluation suites, benchmarks, or A/B testing
- Strong grasp of concurrency, caching, and performance tuning
Nice to have
- Contributions to open-source agent frameworks or AI tools
- Experience with RAG pipelines or vector databases
- Familiarity with Model Context Protocol (MCP)
- Knowledge of LLM inference infrastructure and routing
- Experience with Kubernetes, Docker, or cloud infrastructure
- Background in distributed tracing and production reliability
Stack
- Python
- TypeScript
- LangGraph
- OpenAI Agents SDK
- CrewAI
- AutoGen
- LangChain
- LlamaIndex
- DSPy
- Kubernetes
- Docker