Agentic Engineer (part-time)

Lasker.ai is developing an intelligent AI assistant for Amazon sellers — helping them manage products, reduce account risks, and make smarter business decisions.Powered by official data and advanced AI analysis, our assistant delivers clear guidance, personalized recommendations, and visual insights to improve listings, pricing, and overall performance on Amazon.

We're looking for a staff-level software engineer to help us design & build the next generation of our core agent infrastructure. This is a leadership role.

What You’ll Do

• Own the architecture & continuous improvement of core parts of our agent’s technical infrastructure
• Identify, design, build, ship, and improve new agent features
• Become a world expert at context engineering, including following the latest research & tech. Do whatever is needed to keep us at the frontier.
• Define abstractions that enable the entire team to ship rapidly and reliably on top of your code
• Step up to solve challenges wherever needed—this is a startup, and we're all in it together!

Our ideal candidate

• Very strong CS fundamentals & a systems engineering or distributed systems background
• Exceptional, AI-native programming skills using an ultra-modern stack
• You were the technical lead for a large scale cloud system that’s currently in production
• You’re AGI-pilled & deep into relevant AI tech. You’re currently building an AI agent system.
• You could easily be an engineering leader on an agent product at a top AI lab — but you want the outsized impact of a startup and you believe Lasker can win.

Technical Requirements

• Agentic Orchestration: Deep experience with frameworks like LangGraph, CrewAI, AutoGPT, or PydanticAI. You know how to manage state, memory, and "loop" logic in LLM calls.

• Infrastructure & Tooling: Proficiency in Python/TypeScript and experience with Vector Databases (Pinecone, Weaviate) for RAG-based agent memory.

• The "Human-in-the-Loop" (HITL) Design: Knowledge of how to build robust UI/UX for monitoring agents so humans can intervene before an agent spends $500 on API calls or emails a hallucinated discount to a lead.

• Evaluation & Observability: Experience with tools like LangSmith or Arize Phoenix to track agent performance, latency, and cost-efficiency.

✨ Join us at Lasker.ai — and help build automation that thinks.