Overview
Skills
Job Details
Quick Stats
Rate: $/hr on W2 (No benefits)
Location: 100% On-site (Baltimore/Woodlawn, MD – Leidos Office)
Duration: 3 Years remaining on contract
Clearance: Must obtain Public Trust (4–8 week processing time)
Core Mission
Develop and maintain the Python backend for a public-facing AI Chatbot for the Social Security Administration (SSA). This is a high-availability, production-grade role focusing on RAG, AI Agents, and LLM Safety.
Key Technical Requirements
Python Mastery: 10+ years (or 7+ with Master's) building and scaling production-grade applications. Expert in API design and asynchronous programming.
AI/LLM Production: Hands-on experience with Prompt Engineering, Context Management, and AI Agents.
RAG & Vector DBs: Proficiency in querying Vector Databases for semantic search and implementing Retrieval-Augmented Generation.
LLM Evaluation: Ability to measure performance using frameworks like Ragas, TruLens, or Arize Phoenix (faithfulness, relevance, and safety).
Database Management: Experience with both Relational (SQL) and Vector data structures.
Behavioral & Professional Expectations
Independence: Ability to make architectural decisions in ambiguous environments.
Proactivity: "Startup-like" mindset—vocal, engaged, and highly collaborative.
Engineering Standards: Lead code reviews, enforce CI/CD best practices, and implement robust monitoring/logging.
Primary Deliverables
Backend Engineering: Scalable Python systems for high-traffic, public-facing AI.
LLM Integration: Seamless tool/function calling and structured outputs.
Safety & Compliance: Implementing Guardrails, moderation, and abuse prevention for a regulated government environment.