Position Summary
USP is building a set of AI agents that turn curated Pharmaceutical Supply Chain Data into analyst-ready outputs: Drug Case Study Decks, account email drafts and interactive data lookups. The tools are working for a small team. We need an engineer to make them reliable, secure and self-service for users across the organization. You'll own the path from the data layer to the browser: the agent backend, the web app, hosting, sign-on, and the monitoring that lets people trust the results.
Key Responsibilities:
- Designs module boundaries and contracts across several teams' code.
- Sets the Front-End Architecture and Decides Whether to Migrate Frameworks.
- Designs the Multi-User Hosting Architecture: Scaling, Isolation, High Availability, Cost.
- Designs evaluation, Guardrails and Prompt-Injection Defenses for Agentic Systems.
- Threat Modeling, Leading Security Reviews, Auditability Design.
- Owns the platform roadmap, mentors others, sets standards.
Basic Qualifications:
- 7–10+ years, including technical lead or architect work.
- Performance tuning of Spark on large datasets; data contract design.
- Pharmaceutical, life sciences or supply chain domain experience (NDCs, API manufacturers, drug shortages).
- Working in regulated environments: audit trails, validation, change control (e.g. 21 CFR Part 11 awareness).
- Headless browser automation (Playwright) and document generation (HTML-to-PDF, PPTX)
- Ansible, Bitbucket Pipelines or Jenkins; Nginx.
- Kedro or similar Data Pipeline frameworks; MLflow.
- Agent orchestration patterns (tool calling, bounded loops, LLM-as-judge review).