Senior Multi-Agent LangGraph Engineer (LLM, RAG, Python) FT - Europe, LATAM, NA, Asia

About the job:

We are seeking a Senior Langroid/LangGraph Engineer to lead the design, implementation, and integration of Langroid-based agentic workflows for generative medical reports. This is an advanced, highly specialized role for those with hands-on experience building full-stack LLM systems using Langroid and vector databases such as Qdrant or Chroma. Candidates must be fluent in designing modular, multi-agent systems with robust memory handling, web search tools, citations, and database integrations.

Requirements:

• Extensive hands-on experience building and deploying at least two Langroid (v0.45+) multi-agent systems, including PlannerAgent, SpecialistAgent, and ConsultantAgent, with tailor-plan workflows, checkpointing, metadata-aware orchestration, and citation-aware report generation
• Deep familiarity with Langroid frameworks; experience designing autonomous agent systems with memory routing, inter-agent communication, and modular prompt refactoring
• Experience in Bedrock, backend development, NLP pipelines, or AI system architecture
• Strong Python skills
• Proficient with major LLM APIs including LiteLLM, OpenAI (GPT-4o), Google (Gemini), AWS Bedrock, Anthropic (Claude), and Exa
• Skilled in managing API tokens, context limits, retries, and rate-limiting logic
• Expertise with vector store design and retrieval logic using Qdrant, ChromaDB, Redis, and Pinecone
• Skilled in integrating patient intake forms and user-defined health concerns into prompt generation logic, with support for custom temperature/language settings
• Strong experience with embedding generation (OpenAI, Cohere, HuggingFace) and metadata-filtered searches
• Experience building and tuning retrieval-augmented generation (RAG) systems for latency, cost, citation accuracy, and memory efficiency
• Familiarity with LLM-based PDF parsing libraries (e.g., Marker, GeminiPdfParser), including biomarker extraction, and pipeline comparison (Docling, Unstructured, etc.)
• Advanced LaTeX troubleshooting skills, including citation formatting, patient data referencing, table generation, and error recovery using LatexDebugAgent
• Skilled in Docker containerization, GCP Cloud Run deployment, secure API key management, and .env-based environment setup
• Familiar with lightweight memory profiling, multiprocessing, and cost optimization at scale
  Able to produce clear architectural diagrams, agent workflows, and system documentation
• Experience presenting AI pipeline logic and edge-case handling through code, visuals, or dashboards

Ideal to Also have:

In-depth knowledge of LOINC mapping, biomarker normalization, health-area assignment, and longitudinal lab data synthesis using AI-driven categorization tools
Familiarity with healthcare data standards and regulations (HIPAA, GDPR)
Experience writing unit tests for agents, workflows, and PDF parser configurations
Able to trace full CHR runs in production, debug logs, and resolve misattribution or patient data leakage