🛠 Tech Stack

💼 About This Role

You'll architect and ship the next generation of agentic intelligence across Upwork, leading end-to-end development of AI agents. Your core impact is building a robust platform that enables internal teams and external developers to create reliable, safe, high-performing agents. This hands-on role sits at the intersection of applied research and platform engineering with high visibility across the company.

🎯 What You'll Do

• Design and implement multi-agent systems with guardrails and recovery strategies
• Lead data strategy and LLM alignment (SFT, DPO, RLHF) for agent tasks
• Architect low-latency inference, retrieval, and orchestration services
• Provide technical leadership and mentor senior ICs across teams

📋 Requirements

• 8–12+ years in applied ML/ML systems with 4+ years building LLM-powered products
• Hands-on mastery of LLM adaptation (prompting, tool/function calling) and safety guardrails
• Strong software fundamentals in distributed systems and experience with high-throughput microservices/APIs/SDKs
• Fluency with Python; experience with container orchestration, messaging/streaming, and observability

✨ Nice to Have

• Proficiency in Go, Java, or Javascript
• Experience designing eval suites for agents (task/rubric-based, offline/online)
• Familiarity with agent frameworks and protocols (e.g., MCP)

🎁 Benefits & Perks

• 🏥 Health Insurance
• 🏖️ Unlimited PTO
• 📈 Equity
• 💰 401(k) Matching
• 🧠 Learning Budget