Full-Stack Developer — Business Operating System with Embedded AI Assistant

Full-Stack Developer — Business Operating System with Embedded AI Assistant Project type: Long-term, multi-phase Engagement: Contract, ongoing (potential for ongoing partnership) Hours: Negotiable based on candidate availability Start date: Within 2–4 weeks Location: Remote (US time zone overlap preferred) About the Project We're building a business operating system designed for leadership teams, with one key differentiator: an embedded AI assistant that acts as a chief of staff to every person in a leadership seat. The platform organizes the work of running an organization across six operational areas: Vision, People, Data, Issues, Process, and execution rhythm. The AI assistant draws on the organization's specific context (values, roles, priorities, history) to help leaders think more clearly, prepare better, and act with appropriate judgment. This is not a chatbot bolted onto a CRUD app. The AI is architecturally embedded — sharing the database, authentication, and permission system of the platform itself. Anchor subscriber: A senior care company in Northern California using the product as the first live instance. The platform is designed to be vertical-agnostic and will eventually serve organizations across any industry. What We're Building Phase 1 (current focus) delivers: • Multi-tenant platform with the first organization configured as the anchor subscriber • Backend API supporting six operational components, RBAC, instance isolation • Frontend UI for each component (responsive web, no native mobile apps in Phase 1) • AI assistant with four initial capabilities: conversational data entry, basic retrieval, setup clarification, and priority-to-task decomposition • Authentication and multi-tenant isolation • Full deployment-ready infrastructure The architecture parallels patterns already proven in a sister product (a care management platform built by the same team). Pattern reuse is encouraged: multi-tenant isolation, LLM integration patterns, clarifying-question UX, and task decomposition from a strategic source. What We're Looking For A full-stack developer who is comfortable owning either of two work tracks: Track A — Platform • Multi-tenant database design • Backend API development • Authentication, RBAC, instance isolation • Frontend implementation for six operational components • Instance configuration and data loading Track B — AI Assistant • LLM integration and prompt assembly pipeline • Retrieval architecture (Phase 1 is direct DB queries; later phases add vector retrieval) • Conversation state management • AI panel UI and interaction patterns • Priority-to-task decomposition flow You don't need to commit to a track in your application — we'll discuss fit during the interview. But signal which feels more natural to you. Required Skills • 5+ years full-stack experience with modern web stacks • Strong frontend skills with a contemporary framework (React, Vue, or similar) • Strong backend skills with REST API design, relational database modeling, authentication, RBAC • Multi-tenant architecture experience — strict instance isolation is a non-negotiable requirement • Mobile-responsive design experience (we're not building native apps in Phase 1) • Clean, documented, testable code — this is a long-term codebase, not a prototype • Experience working from product specifications and architectural documents (we have substantial foundation documentation) Preferred Skills • LLM integration experience — prompt engineering, retrieval pipelines, conversation state, provenance logging • Experience with conversational UX — chat interfaces that aren't generic chatbots • Healthcare or compliance-adjacent experience (HIPAA awareness, data isolation, audit logging) • Experience extending existing codebases with shared patterns from related products • AI/ML background — embedding stores, vector retrieval, semantic search (relevant for Phase 2+) What You Won't Be Doing To be honest about scope: • This is not a greenfield AI research project. We're integrating LLMs, not training models. • This is not a quick MVP. We're building a real product with multi-phase scope. • This is not a project where you'll be making product strategy decisions alone — the founder is deeply involved in product direction. • This is not a customer-support chatbot. The AI assistant is a chief-of-staff partner, not a Q&A bot. What's Already in Place You won't be starting from a blank brief. We have: • Comprehensive foundation documentation — purpose, architecture, knowledge architecture, Phase 1 development brief, AI system prompt, sample interactions (these will be shared with serious candidates after initial conversation) • Decided interaction model — UI layout and AI panel placement already specified • Designer in place — full-time UI/UX designer producing wireframes and visual system in parallel • Domain expertise — the founder runs the anchor subscriber's business and brings deep operational knowledge • Sister product to reference — many architectural patterns have been built and tested in a related product; reuse is encouraged You'll be collaborating with: one other developer (specialized in the complementary track), a full-time UI/UX designer, and the founder (acting as product owner). How We Work • Weekly syncs (30–45 minutes) with the full team • Milestone demos every 2–3 weeks • Async communication via [your preferred tool — Slack, etc.] for day-to-day questions • Iterative scope refinement — we adjust as we learn, but we protect Phase 1 scope from drift • Founder is accessible — typical async response within 24 hours What to Include in Your Application 1. Brief introduction — who you are and why this project interests you 2. Track preference — Track A (Platform) or Track B (AI Assistant), with reasoning 3. Relevant work examples — 2–3 projects you've shipped that demonstrate the required skills (especially multi-tenant work or LLM integration) 4. Your typical availability — hours per week and time zone 5. Your hourly rate — please be specific Please skip: • Generic "I have 10 years of experience" cover letters • Lists of every technology you've ever touched • Boilerplate Upwork applications Selection Process 1. Application review (1 week) 2. Short async screening (a few questions about architecture and approach) 3. 30-minute video conversation with the founder 4. Paid trial task 5. Offer Compensation Hourly contract at market rates for experienced full-stack developers with the listed skills. Specific rate negotiable based on experience and track. Long-term engagement preferred — this is a multi-phase build, not a single sprint. A Note on Quality We're building something that will be used by real leaders running real organizations. Code quality, AI behavior quality, and design quality all matter. We're looking for someone who cares about craft — who reads documentation before asking questions, ships working demos rather than describing planned features, and treats the AI assistant's voice as a product requirement, not an implementation detail. If that resonates, we'd like to hear from you.