From Prompt to Plan: Agentic Interfaces for Human-in-the-Loop Operations Research
Scope
Build an agentic orchestration layer that translates natural‑language prompts into solver actions, learns user preferences, and enables near‑real‑time, planner‑in‑the‑loop re‑optimization for complex OR problems.
What you’ll do
Prototype parser agents, wire up solver APIs, design preference‑learning feedback loops, and optimize latency/UX for trustworthy, explainable outputs.
Profile
- Excellent Python
- Enthusiasm for mathematical optimization (MILP/CP/SAT)
- Familiarity with modern LLM tooling (function‑calling / agent frameworks)
- Bonus: basic DevOps or front‑end
- Curiosity, initiative, and collaborative mindset
Details
- Zurich‑based, mostly in‑person preferred.
- Start: As soon as possible
- Co‑supervised with Prof. Menna El‑Assady and IVIA Lab @ ETH Zurich
Apply (ETH Zurich only)