About the Project We are building an AI-powered Patent Search & Analysis Platform that will assist law firms, corporations, and inventors in conducting novelty (patentability) searches efficiently. The platform will leverage multiple Large Language Models (LLMs) along with structured patent databases to deliver high-quality, human-like patent search reports. Scope of Work The selected freelancer will be responsible for designing and developing the core AI system and backend workflow, including: Building an AI-driven patent search pipeline Integrating multiple LLM APIs (e.g., OpenAI, Claude, etc.) Developing logic for: Query understanding (invention → search strategy) Keyword + semantic search Prior art extraction Claim comparison Structuring outputs into readable patent search reports Integrating patent databases/APIs (e.g., Google Patents, Lens.org, or paid databases if needed) Creating a scalable backend architecture Key Responsibilities Convert invention disclosures into structured search queries Implement semantic search + vector embeddings Design multi-step AI workflows (agent-based or pipeline-based) Build ranking and filtering mechanisms for prior art Generate automated patentability reports Ensure accuracy, relevance, and explainability of results Optimize the cost efficiency of LLM/API usage Technical Requirements Must-Have Skills Strong experience in Python Experience with LLM integration (OpenAI, Anthropic, etc.) Knowledge of: Vector databases (FAISS, Pinecone, Weaviate, etc.) Embeddings & semantic search Experience with API integrations Familiarity with backend frameworks (FastAPI, Flask, etc.) Good-to-Have Experience with patent data or legal tech Knowledge of RAG (Retrieval-Augmented Generation) architectures Experience with LangChain / LlamaIndex or similar frameworks Understanding of patent search methodologies (novelty/FTO/validity) Deliverables Fully functional AI patent search backend system Modular architecture for: Query processing Search execution Result ranking Report generation API endpoints for frontend integration Basic documentation of system workflow Budget Open to proposals (Fixed / Milestone-based) Indicative budget aligned with complexity (mid to high range) What We’re Looking For Someone who can think like a product builder, not just a coder Ability to translate business requirements into AI workflows Strong problem-solving and communication skills Prior experience in AI SaaS products is a big plus To Apply, Please Include Links to relevant AI/LLM projects Brief explanation of how you would approach: “Building an AI-based patent search system” Your suggested tech stack & architecture Estimated timeline and cost