AI/LLM Full-Stack Engineer to add AI/LLM features to an existing enterprise SaaS platform

We run an established, enterprise-scale SaaS platform that has been in production for many years with a large active user base. We are now investing in a major modernization initiative: integrating AI/LLM capabilities (Claude, OpenAI/GPT-4, and similar models) directly into our existing product to make it smarter, more automated, and more user-friendly. This is not a greenfield prototype — you'll be working inside a mature, production codebase, so we need someone who has done real-world AI integration into existing systems, not just demos or tutorials. Tech Stack ● Frontend: Next.js / React / TypeScript ● Backend: Node.js and Python (REST/GraphQL APIs) ● Infrastructure: Enterprise-scale, cloud-hosted, multi-tenant What You'll Be Doing ● Design and implement AI/LLM features into our existing SaaS workflows ● Integrate Claude API (Anthropic) or OpenAI APIs, with an architecture flexible enough to support multiple/model-agnostic providers ● Build RAG pipelines with vector databases so AI responses are grounded in our platform's data ● Implement prompt engineering, function/tool calling, streaming responses, and agentic workflows where appropriate ● Ensure enterprise-grade concerns are handled: data privacy, multi-tenancy isolation, rate limiting, token/cost optimization, latency, observability, and fallback handling ● Build polished, user-friendly AI UX in Next.js ● Collaborate with our existing team, follow our code standards, write clean, tested, maintainable code Required Experience ● 5+ years full-stack development (Next.js/React + Node.js and/or Python) ● Proven, shipped experience integrating LLMs (Claude/Anthropic API, OpenAI API, Azure OpenAI, or open-source models) into existing production systems ● Hands-on with RAG, embeddings, and vector databases ● Strong API design and system architecture skills at enterprise scale ● Understanding of AI security, PII handling, and cost/token management ● Excellent communication in English; able to propose solutions, not just take tickets Nice to Have ● LangChain / LlamaIndex / Vercel AI SDK experience ● Fine-tuning or model evaluation experience ● Experience with AWS Bedrock / GCP Vertex AI ● Prior work on B2B/enterprise SaaS products Engagement Long-term potential for the right person. We'll start with a well-defined initial milestone (paid), and expand from there. Please include links to AI features you've personally shipped in production. To filter out generic/AI-generated proposals, start your proposal with the word "Falcon" and briefly describe the most complex LLM integration you've built.