Senior AI Developer

As a Senior AI Developer at K3-Innovations, Inc., you will play a crucial role in designing and optimizing retrieval-augmented generation (RAG) pipelines for knowledge workflows in the biopharma industry. Your responsibilities will involve combining expertise in database design, vector search optimization, backend architecture, and LLM integration to create a scalable AI platform that connects structured and unstructured biopharma data with advanced AI technologies. - **RAG Pipeline Design and Optimization:** - Architect and implement retrieval-augmented generation pipelines integrating document retrieval and LLM response generation. - Design and maintain knowledge bases and vector stores using tools like FAISS, Weaviate, or PostgreSQL PGVector. - Optimize retrieval mechanisms (chunking, indexing strategies, reranking) to enhance response accuracy and efficiency. - Integrate context-aware querying from structured (Postgres) and unstructured (text/PDF) sources. - **Database and Embedding Management:** - Design relational schemas to support knowledge base metadata and chunk-level indexing. - Manage embeddings pipelines using open-source models (e.g., HuggingFace sentence transformers) or custom embedding services. - Optimize large-scale vector search performance (indexing, sharding, partitioning). - **LLM and Prompt Engineering:** - Develop prompt engineering strategies for retrieval-augmented LLM pipelines. - Experiment with prompt chaining, memory-augmented generation, and adaptive prompting techniques. - Fine-tune lightweight LLMs or integrate APIs from OpenAI, Anthropic, or open-source models (e.g., LlamaIndex, LangChain). - **Backend API and Workflow Orchestration:** - Build scalable, secure backend services (FastAPI/Flask) to serve RAG outputs to applications. - Design orchestration workflows integrating retrieval, generation, reranking, and response streaming. - Implement system monitoring for LLM-based applications using observability tools (Prometheus, OpenTelemetry). - **Collaboration and Platform Ownership:** - Work closely with platform architects, AI scientists, and domain experts to advance knowledge workflows. - Take ownership from system design to model integration and continuous improvement of RAG performance. **Qualification Required:** - AI RAG Engineering: - Experience building RAG architectures in production environments. - Expertise with vector stores (e.g., FAISS, Weaviate, Pinecone, PGVector). - Experience with embedding models and retrieval optimization strategies. - Prompt Engineering: - Deep understanding of prompt construction for factuality, context augmentation, and reasoning. - Familiarity with frameworks like LangChain, LlamaIndex, or Haystack. - Database and Backend Development: - Strong proficiency in relational and vector extension design (PGVector preferred). - SQL optimization, indexing strategies for large datasets. - Experience building backend services using FastAPI or Flask. If you are passionate about building intelligent retrieval systems, fine-tuning prompt pipelines, and optimizing LLM-based applications for real-world datasets, this role at K3-Innovations, Inc. offers you the opportunity to have a meaningful impact, deepen your expertise, and contribute to a results-driven work culture with remote flexibility and competitive compensation. Join the team at K3-Innovations and be part of shaping the future of AI in the biopharma space.,