Description
Summary We are seeking an established engineering company with proven experience in AI-native architecture and scalable multi-tenant SaaS infrastructure to deliver the MVP build of Hospitality B2B operational platform. The architectural foundations, Architecture Decision Records (ADRs), and the full UI/UX design system are already locked. We are now selecting the team that will implement the roadmap. This is a fixed-scope, fixed-timeline engagement with a clear MVP boundary and a documented set of architectural constraints. We are not looking for an agency to define the architecture or design the product — that work is complete. We are looking for a senior, well-organised team that can execute against the spec and deliver to the deadline. What's already in place: Production-grade Sprint 0 architectural specification: multi-tenant, multi-region, AI-native ADRs covering every architecturally significant decision Locked technology stack: PostgreSQL with Row-Level Security, pgvector with tenant-scoped namespace partitioning, event-driven substrate (AWS EventBridge), AI decision broker with five-tier routing policy Complete UI/UX design system covering 4 user surfaces: operator web app, worker mobile PWA, guest portal and read-only client portal Compliance posture designed in from Day 1: ISO 42001 (AI governance), SOC 2 Type II, and ISO 27001 ISMS architecture — phased build with external certification sequenced 12–18 months post-launch An on-retainer Architect of Record who will validate architectural conformance throughout the build and approve any proposed deviations from the locked ADRs MVP scope boundary clearly defined; deferred features are explicitly trigger-gated, not calendar-gated What we need from your team: Execution of the MVP build against the locked architecture and design system Implementation of four external system integrations conforming to a canonical internal data model Multi-agent orchestration substrate built around generic agent contracts (substrate must support future specialist agents as configuration work, not platform work) Retrieval-Augmented Generation (RAG) pipelines with tenant-scoped vector storage Robust DevOps from Day 1: CI/CD, environment topology, observability, SLO instrumentation, append-only audit log Bi-weekly architectural validation cycles with the Architect of Record Written conformance to ADRs; any proposed deviation requires written approval Required experience: Established company with documented org structure, named delivery leads, and demonstrable continuity of team — not a freelance collective Proven track record building AI-native production systems (LLM orchestration, RAG, agent architectures) — please reference specific shipped systems in your proposal Multi-tenant SaaS with Row-Level Security at scale Microservices and event-driven architecture patterns Strong front-end capability: React, React Native or PWA, and implementation against a provided design system Demonstrated ability to deliver to fixed milestones and timelines Senior engineers are comfortable working under architectural constraints set by an external Architect of Record Bonus experience (any of the following will strengthen your proposal): PropTech, HospitalityTech, multi-site operations platforms, or distributed asset management systems Workforce scheduling, dispatching, or field operations systems Compliance-led builds (ISO 27001, SOC 2 Type II, or ISO 42001) Vector-store interface abstractions enabling future pgvector → managed vector DB migration without application-layer rewrites Engagement model: Fixed-price milestone-based engagement via Upwork escrow Kickoff anchor: early July 2026 MVP delivery target: late 2026 / early 2027 NDA required before architecture document or design system is shared Architectural deviations from the ADR set require written approval from the Architect of Record (this will be a contractual obligation) Next steps: Submit a proposal including: Company background and org structure Three relevant case studies with measurable outcomes Proposed delivery team (named roles, named engineers where possible) High-level cost estimate and proposed milestone schedule Any clarifying questions before NDA signing Shortlisted companies will be invited to a 45-minute technical screening call with the Architect of Record. Final selection criteria: architectural fluency, delivery track record under fixed milestones, and demonstrated ability to work within external architectural constraints. We are not looking to be the first AI project on a company's portfolio. If your team has not shipped multi-tenant AI-native production systems before, please save us both the time.