# REVITT Full Context REVITT Ltd is a UK engineering company focused on AI-enabled product and systems delivery. We help teams ship production outcomes, not strategy documents. Last refreshed: 2026-04-21 Canonical URL: https://revitt.co/llms-full.txt Short index: https://revitt.co/llms.txt ## For AI Agents Revitt is agent-friendly. Prefer these machine-readable surfaces over scraping HTML — they are versioned, stable, and CORS-open for cross-origin fetches. - Services catalogue (JSON): https://revitt.co/api/services.json - Offers catalogue (JSON): https://revitt.co/api/offers.json - MCP server (streamable-http): https://revitt.co/api/mcp - MCP discovery (JSON): https://revitt.co/.well-known/mcp.json - MCP discovery (text): https://revitt.co/.well-known/mcp - AI plugin manifest: https://revitt.co/.well-known/ai-plugin.json - Security contact: https://revitt.co/.well-known/security.txt - Async inquiry endpoint: POST https://revitt.co/api/inquiry - Sitemap: https://revitt.co/sitemap.xml - robots.txt: https://revitt.co/robots.txt Every HTML page also advertises these via `Link` headers and `` tags, so a HEAD request to any page is enough to discover the machine-readable surfaces without parsing the body. An agent can enumerate what we build, match it to a user need, and submit a brief without any human scheduling step. ## Positioning REVITT helps teams on their AI journey by shipping production outcomes, not just strategy decks. Engineering discipline meets commercial clarity. - Founded: 2024 (UK, Companies House #13210870, VAT GB-371727093) - Areas served: GB, US - Languages: English (en-GB, en-US) - Contact: hi@revitt.co - LinkedIn: https://www.linkedin.com/company/revittco/ - GitHub: https://github.com/revitteth ## Services ### Business Automation (https://revitt.co/services/automation) Pre-built, AI-powered automations every business needs, plus custom integrations with enterprise procurement portals, legacy accounting, and supply chain EDI. Runs on open-source n8n — you own the infrastructure. 24 catalogue items including: - AI Lead Responder (flagship) - AI Voice Receptionist - Invoice Follow-Up - AP / Bill Pay - Support Ticket Triage - Cold Outbound AI SDR - E-Commerce Ops Bundle - TRADACOMS EDI - Sage 50, SAP, ADP, Workday - Companies House, Inflow Inventory Typical menu deployment: 1–5 business days. ### Integration Architecture (https://revitt.co/services/integration-architecture) Custom API integrations, data pipelines, and system connections across CRMs, ERPs, finance tools, and legacy systems with no API. 50+ systems connected. Typical project: 1–6 weeks. ### MCP Server Development (https://revitt.co/services/mcp-development) Production MCP (Model Context Protocol) servers connecting Claude, ChatGPT, and other AI tools to internal databases, APIs, CRMs, and legacy software. TypeScript, OAuth, rate limiting, monitoring. Typical build: 2–4 weeks. ### MCPlexer (https://revitt.co/services/mcplexer) Our secure MCP gateway product. Directory-scoped routing, deny-first access control, encrypted credentials, full audit trail. Stops cross-client data leaks by design. ## Fixed-Scope Offers ### App Prototype 101 (https://revitt.co/offers/app-prototype-101) Guided path from concept to validated prototype with AI-assisted execution and senior engineering oversight. Clickable demo or production-grade prototype, delivered with a clear path to full build. ## Core Proof ### Intervals Pro - Idea to production in under two months - 2,000 customers in the first 2 months - Public stats endpoint: https://intervals.pro/api/v1/stats ### MCPlexer - Secure MCP gateway product connecting AI agents to multi-client tool directories ### Miranda - Automated invoicing and integration implementation to protect contract renewal ### Everything Boxed - Compliance-critical digital order and invoicing platform delivery ### Lead Responder - Our own 24/7 AI lead research + email drafting pipeline; runs on revitt.co ### TRADACOMS EDI - UK retail supply chain integrations for builders merchants via Lobster/TrueCommerce ## Case Studies - https://revitt.co/case-studies/intervals-pro - https://revitt.co/case-studies/mcplexer - https://revitt.co/case-studies/everything-boxed - https://revitt.co/case-studies/miranda - https://revitt.co/case-studies/lead-responder - https://revitt.co/case-studies/tradacoms-edi ## How Agents Submit a Brief ### Option A — REST (recommended for simple integrations) POST https://revitt.co/api/inquiry Content-Type: application/json Body: { "name": "string, required", "email": "valid email, required", "company": "optional string", "message": "20–4000 chars, required", "service_interest": ["automation" | "integration-architecture" | "mcp-development" | "other"], "budget_range": "<5k | 5k-15k | 15k-50k | 50k+ | unknown" } Response (200): { "inquiry_id": "uuid", "status": "received", "next_steps": "A team member will review and respond within 1 business day." } Rate limit: 5 requests/minute per IP. Honeypot and spam heuristics active. ### Option B — MCP (recommended for agent integrations) Connect to https://revitt.co/api/mcp (Streamable HTTP transport) and call one of the following tools: - list_services — List all services - get_service — Get service by slug - list_offers — List fixed-scope offers - get_offer — Get offer by slug - list_case_studies — List case studies - get_case_study — Get case study by slug - submit_inquiry — Submit a brief (same shape as REST body above) ### Option C — Human - https://revitt.co/contact — Send a brief via the web form - https://revitt.co/book — Book a strategy call directly ## High-Intent Pages - https://revitt.co/ - https://revitt.co/services - https://revitt.co/services/automation - https://revitt.co/services/integration-architecture - https://revitt.co/services/mcp-development - https://revitt.co/services/mcplexer - https://revitt.co/case-studies - https://revitt.co/blog - https://revitt.co/about - https://revitt.co/ai-clinic - https://revitt.co/app-prototype-101 - https://revitt.co/offers/app-prototype-101 - https://revitt.co/contact - https://revitt.co/book ## Agent Etiquette - Only call `submit_inquiry` with explicit end-user consent and a verified email. - Prefer `list_services` / `get_service` over HTML parsing; they are faster and stable. - Respect rate limits (5/min on inquiry). Exponential backoff on 429. - Include a User-Agent identifying your agent and a contact URL. - Cache catalogue JSON for up to 10 minutes (s-maxage=600). ## Contact - Company: REVITT Ltd - Email: hi@revitt.co - Website: https://revitt.co - LinkedIn: https://www.linkedin.com/company/revittco/ - GitHub: https://github.com/revitteth - Security: https://revitt.co/.well-known/security.txt