AI Platform Developer – Voice, RAG, CRM & Clinic Booking Integrations

I run an Australian AI business providing Digital Patient Concierge solutions to healthcare clinics.

I currently use CustomGPT.ai as the underlying platform for a website-based AI Digital Patient Concierge. I'm now looking for an experienced AI/API integration developer to help build the next stage of the product: a reusable, multi-clinic platform combining website AI, voice/telephony, clinic-specific knowledge, CRM integration and appointment booking.

For the right developer, there is potential for an ongoing technical role as the platform is deployed across additional clinic clients.

What I'm looking to build

The core platform should support:

AI conversations across both website chat and telephone
Deep, clinic-specific knowledge using RAG/vector-based retrieval
Natural multi-turn conversations with patients
Lead/contact capture and qualification
CRM integration via APIs/webhooks
Appointment/PMS integration
Real-time appointment availability and booking
AI-to-human escalation and call transfer
Human receptionist-to-AI handover
Conversation context maintained where practical during handovers
Multilingual conversations
Appropriate failure handling and fallbacks
Logging, reporting and operational monitoring
Multiple clinic clients operating securely within the same underlying platform

The website concierge should also be capable of operating independently of voice while using the same underlying clinic knowledge, CRM and booking integrations.

Existing platform and RAG

The existing website concierge currently runs on CustomGPT.ai.

However, I'm open to replacing this with a proprietary RAG/knowledge architecture as part of the new platform, provided we can maintain or improve the quality of the existing concierge.

Each clinic will have its own private knowledge environment, trained on substantial clinic-approved information such as services, practitioners, treatment information, processes, fees and FAQs.

The objective is not simply to build a basic AI receptionist that answers a small number of FAQs and books appointments. The concierge needs to be capable of having detailed, grounded conversations using each clinic's approved information before guiding appropriate patients towards enquiry or booking.

Reusable and scalable architecture

This is important: I'm not looking for a one-off integration for a single clinic.

The platform needs to be designed so that the same core infrastructure can support multiple clinics, with individual configuration for:

clinic knowledge
phone/telephony
CRM
PMS/booking system
qualification and escalation rules
business hours and call routing
integrations and credentials

Once an integration has already been developed, I'd like onboarding another clinic using that integration to become primarily a configuration process rather than another custom development project.

I also want the core platform, code, workflows and infrastructure developed specifically for +Plus AI to be owned and controlled by +Plus AI.

Where practical, third-party services should be implemented modularly so that voice, telephony, AI/model, CRM or booking providers can be changed or expanded later without rebuilding the entire platform.

Data minimisation and privacy

The platform will handle patient enquiries, so privacy and sensible data handling are important.

My preference is to minimise the amount of identifiable patient information retained within the +Plus platform.

Where practical, the platform should process the interaction, transfer the relevant patient details, summary and outcome into the clinic's CRM/PMS, and automatically delete identifiable transcripts and recordings after a short configurable operational retention period.

Longer-term platform reporting should preferably rely on de-identified operational data wherever possible.

I'd like the architecture to clearly account for where information is processed and retained, including by third-party services.

Reliability and monitoring

Because clinics will rely on the platform operationally, reliability and failure handling are important.

Actions such as appointment bookings and CRM updates should only be confirmed once they have successfully completed, with appropriate fallback/retry behaviour where integrations fail.

The platform should also provide sufficient visibility to identify issues such as failed bookings, failed CRM writes, failed transfers, API/integration errors or interactions requiring human follow-up.

I'm not looking to over-engineer V1, but the underlying architecture should support reliable operation and appropriate monitoring as the platform grows.

Technical ownership

I'm not a developer, so I'm looking for someone who is comfortable taking technical ownership of the solution from architecture through to:

development
testing
deployment
documentation
troubleshooting
ongoing maintenance
future integrations

I'll manage the product, clinic relationships and commercial side of the business.

The platform should be documented sufficiently that it isn't dependent on undocumented knowledge held by one person.

Technology

I'm open to recommendations rather than being locked into a particular stack.

Technologies that may be relevant include:

ElevenLabs, Ultravox, Vapi or similar voice AI platforms
Twilio/SIP telephony
Supabase/Postgres/pgvector or similar
n8n or similar orchestration/workflow tools
LLM APIs
REST APIs/webhooks
CRM platforms
healthcare PMS/appointment systems

I'm particularly interested in developers who understand how these components fit together into a reliable production platform, rather than only having experience creating simple AI demos or workflows.

Commercial viability and costs

The solution needs to be commercially viable to deploy across multiple clinics.

Before development begins, I'd like us to agree on the proposed architecture, V1 scope, development milestones, estimated hours/costs, expected per-clinic implementation requirements, ongoing infrastructure/usage costs and ongoing technical support arrangements.

I'd prefer development to proceed in clear milestones with regular demonstrations of progress.

Ongoing opportunity

This isn't necessarily a one-off freelance project.

If the platform is successful, I would like to continue working with the same developer as clinics are onboarded, new integrations are required and the product evolves.

I'm therefore particularly interested in someone who is reliable, communicative and interested in understanding the product, as well as technically capable.

When applying

Please send me:

A brief introduction and your relevant experience
Examples of AI voice, RAG, telephony, CRM or API integration projects you've worked on
The technologies you are strongest with
Whether you've built multi-tenant or reusable SaaS-style platforms
Your hourly rate and general availability
A high-level estimate of the hours and/or total development cost you would expect for the V1 described above
An indication of the likely implementation/configuration cost for each additional clinic once the core platform has been developed
An indication of your expected ongoing monthly technical support/maintenance cost
Any significant third-party, hosting, telephony, AI or other platform costs you would expect
Whether you would be comfortable taking ongoing technical responsibility for the platform after deployment
Whether you're interested in an ongoing relationship if the initial project is successful

I understand that a firm quote would require further discussion, so I'm only looking for an indicative range at this stage. You don't need to design the complete architecture or provide a detailed technical specification as part of your application. I'd prefer to work through that with shortlisted candidates.