Software Engineer (Agent Platform) - Intern - 2026-2027

What You'll Do

• Build the orchestration layer: Design and ship the systems that route, sequence, and supervise multi-agent workflows across customer environments in real time.
• Extend agent harnesses: Build the tooling, sandboxes, and eval loops agents use to test and validate their own outputs before changes reach production.
• Work on Agent Manager: Contribute to the internal system where agents propose, test, and roll out improvements to other agents. This includes guardrails, rollback paths, and human-in-the-loop checkpoints.
• Ship agentic products: Design, code, and ship full-stack features for Netic's agent platform, from data models to APIs to front-end tooling.
• Co-create with customers: Work with real customer workflows to surface where automated agent-improvement can and can't be trusted, and turn those edge cases into platform safeguards.

What You'll Learn / Bring

• Systems fundamentals: Comfort with Python (asyncio, FastAPI) and TypeScript (Node, React); CS fundamentals in distributed systems.
• Agent tooling exposure: Coursework, internship, or personal-project experience with LLM APIs, eval frameworks, RAG, or agent frameworks (LangChain, MCP, or similar) — production experience not required, curiosity is.
• Orchestration mindset: Interest in how multi-step, multi-agent systems coordinate state, tool calls, and failure recovery — you don't need to have built this before, but you should be excited to.
• Founder-level ownership: Track record of shipping full projects (school, internship, or personal) end-to-end, not just isolated assignments.
• Comfortable with ambiguity: You'll be working on bleeding edge areas within agentic AI, but that means you'll be tackling problems that don't have established best practices yet. There isn't much of a playbook for agents building agents. You should be excited to help define them and write it.

What Brings Us Together

• Our Commitment To Live to build
• Run through walls and win
• Obsess over customers in each line of code
• Lose sleep over the "almost perfect"
• Show internal locus of control
• Prioritize finesse: refinement of first principles thinking, execution, and craftsmanship