Senior Python / LLM Engineer Needed – Router MVP & Predictive Model Loader

✅ In-Scope Work (Remaining) Milestone 1 Router MVP Implementation Deliverables • Embedding pipeline (query + document embeddings) • Vector storage using FAISS or Chroma • Projection module for query feature extraction • Configurable scoring & model-selection strategies • Router MVP with pluggable LLM backends • Router validation tests (routing correctness & mis-routing analysis) Acceptance Criteria • End-to-end routing demonstrated • Deterministic and explainable routing decisions • Pytest unit tests included • All code committed to repository ⸻ Milestone 2 Predictive Loader & Integration Deliverables • Predictive model loader (LLM-based classifier) • Warm-start caching & preload logic • Cache management strategy • Full integration with Router MVP • FastAPI backend exposing routing endpoints • Structured JSON logging • End-to-end testing + documented stress testing • Final validation & testing report Acceptance Criteria • Predictive loading working correctly • Fully integrated end-to-end system • Stress-testing methodology clearly documented • Final report delivered ⸻ Technical Requirements • Strong Python backend experience • FastAPI • LangChain or equivalent LLM orchestration framework • FAISS or Chroma vector stores • Dockerized services • Structured JSON logging • Pytest for unit & integration testing ⸻ Ideal Candidate • 3+ years of Python backend experience • Proven experience with LLMs, embeddings, and routing systems • Hands-on with vector databases and retrieval pipelines • Comfortable writing clean, testable, production-ready code • Experience with performance testing and system validation • Strong communication and documentation skills ⸻ Deliverables & Collaboration • All work delivered via Git repository • Clean, readable, well-tested code • Clear documentation for setup, testing, and usage • Milestone-based payments ⸻ To Apply Please include: 1. Relevant experience with LLM routing, embeddings, or RAG systems 2. GitHub or code samples (if available) 3. Brief explanation of how you would approach Router MVP + predictive loading Apply Now Apply Now