Hybrid AI Pipeline Development: Design and deploy advanced GenAI architectures leveraging a mix of commercial APIs and locally hosted open-source models optimized for low latency and data containment.
ERP Framework Integration: Develop, optimize, and maintain custom Python backend modules to seamlessly bridge complex LLM logic with Odoo’s native ecosystem, database models, and relational data structures.
Portal-Level System Deployment: Implement secure, scalable conversational layers and intelligent utility interfaces within internal employee portals and client-facing web surfaces.
AI Safety & Alignment Engineering: Build strict input/output verification layers, semantic boundary guardrails, and automated filtering to ensure production-grade safety, data privacy, and deterministic reliability.
Knowledge Retrieval Systems: Construct, benchmark, and maintain advanced Retrieval-Augmented Generation (RAG) pipelines over dynamic, multi-source enterprise knowledge bases.
Core Language & Engineering: Mastery of Python, object-oriented programming, and clean, scalable backend architecture practices.
ERP Ecosystem: Hands-on familiarity or a strong conceptual understanding of Odoo development (Odoo ORM, Python, XML, PostgreSQL).
Local Inference & Deployment: Direct exposure to running local models, model quantization techniques, and inference engines such as Ollama, vLLM, or Hugging Face TGI.
Custom LLM Orchestration: Ability to write raw Python orchestration logic for handling context windows, state management, and direct model-to-tool API execution without relying on heavy third-party agent wrappers.
AI Safety Infrastructure: Familiarity with LLM guardrail frameworks (e.g., NeMo Guardrails, Llama Guard) or custom rule-based validation engineering techniques.
Vector Architectures: Experience utilizing vector databases (e.g., ChromaDB, FAISS, Qdrant, or Pinecone) for high-dimensional similarity search.
