Senior AI / MCP Developer — Multi-Tenant AI Integration Platform (ChatGPT, Claude, APIs)

Project Overview We are seeking a senior AI/API integration developer or architect to design and build Phase 1 of a secure, multi-tenant AI integration platform. This is NOT a website redesign or WordPress development project. Our company website already exists. The purpose of this project is to create an integration layer that allows AI assistants such as ChatGPT and Claude to securely interact with business systems through MCP/API connections. AIrsa Leads will be the first workspace/tenant, but the architecture should be designed from the beginning so additional businesses/workspaces can be added later without rebuilding the platform. Primary Phase 1 Objectives Multi-Tenant SaaS Foundation Secure multi-tenant architecture Separate workspace/tenant configurations Authentication and authorization Secure credential/API-key management Architecture that allows additional companies to be added later Proper logging, error handling and basic monitoring MCP / AI Assistant Integration MCP-compatible integration for ChatGPT Integration with Claude where technically appropriate Clean abstraction so additional AI models/providers can be added later We do NOT want the platform permanently dependent on a single AI provider GoHighLevel Integration / Discovery AIrsa Leads currently uses GoHighLevel as its primary CRM, phone/SMS, automation, scheduling and client communication platform. Phase 1 should determine and document: Which GHL data/actions are accessible through its APIs Authentication requirements API limitations Webhook capabilities Rate limits Which functions can safely be exposed to AI assistants Recommended architecture for future deeper GHL integration Normalized Integration Layer Results from different providers should be returned to the AI layer in a consistent standardized structure, rather than requiring the AI assistant to understand every provider's unique API format. For Phase 1, we are primarily looking for normalized responses/interface behavior. We are NOT currently asking for a large historical data warehouse or full cross-system analytics database unless you recommend that it is necessary. Security Security is extremely important. Architecture should include: Tenant isolation Least-privilege access Secure secrets/API credential storage Authentication/authorization Audit/logging strategy Protection against one tenant accessing another tenant's information Appropriate handling of customer/business data Documentation & Handoff We expect: Architecture documentation Source code Deployment documentation Environment/configuration documentation API/MCP documentation GHL discovery findings Testing Clear handoff so another qualified developer can maintain the system Important Architectural Principle We want the business workflows, integrations and customer data layer to be the durable asset, rather than hard-wiring the business to ChatGPT, Claude or any single AI model. The architecture should make it reasonably straightforward to add or replace AI providers over time. Future Direction — Not Necessarily Phase 1 The architecture should leave room for future integrations with additional systems and additional businesses. AIrsa Leads is the first tenant. We have other businesses that may eventually use separate workspaces, and there is a possibility that the platform could eventually become a commercial SaaS product. We do not want to over-engineer Phase 1, but we also do not want to build a dead-end prototype that must be replaced if the concept succeeds. What We Need From You Please review this scope and include the following in your proposal: Your recommended technical architecture. Technologies/frameworks you would use and why. How you would implement multi-tenancy. Your experience building MCP servers or AI-agent/API integration platforms. Your experience with OpenAI/ChatGPT and Claude APIs. Any experience integrating GoHighLevel. What you believe should and should NOT be included in Phase 1. Estimated development timeline. Fixed-price estimate for Phase 1, preferably broken into milestones. Any ongoing infrastructure/API costs we should expect. What assumptions your estimate is based upon. Examples of comparable systems you have personally built. Please do not simply provide an hourly rate. We specifically want your independent assessment of the scope and a fixed-price estimate. We are looking for someone capable of thinking like a software architect, not simply connecting a few APIs. We value simplicity and maintainability. If part of this scope is unnecessarily complicated or there is a substantially better way to accomplish the objective, we want you to tell us.