Skip to content
← All projects

AI Integration + Full-Stack Delivery

myFelipe.ai

myFelipe.ai turns conversational AI into a practical front door for a business: answering calls around the clock, handling multilingual conversations, and moving callers toward booked appointments instead of missed opportunities.

My role

Backend implementation, systems integration, and production deployment

Core stack

Backend Development · Retell AI · OpenAI · AWS

Visit live project ↗
Responsive desktop presentation of the myFelipe.ai website.
Responsive desktop presentation of the myFelipe.ai website.

Challenge & approach

The challenge

The product needed more than a compelling AI demo. Its customer-facing experience, backend behavior, third-party AI services, and cloud environment all had to work as one dependable production system.

The approach

Harley implemented backend functionality within the client engagement, worked across the integration boundaries that connect the product experience to its AI capabilities, and deployed the finished website to an AWS cloud server. The work demonstrates the ability to move between application logic, external services, infrastructure, and release execution without losing sight of the user journey.

System architecture

Experience layer

A conversion-focused web experience introduces the product, communicates value, and gives prospects a path into the live AI demo.

Application layer

Backend functionality supports the product workflows and keeps customer-facing behavior separate from infrastructure concerns.

AI service boundary

Retell AI and OpenAI capabilities sit behind a defined integration boundary so conversational services can evolve without coupling the entire application to one provider.

Cloud runtime

The production release runs on AWS with the application and deployment configuration treated as part of the delivered system.

Notable work & outcomes

  • Treated AI as a product capability with real user and business workflows, not as a standalone prototype.
  • Worked across backend code, external AI services, hosting, and production release concerns.
  • Owned the last mile from implemented functionality to a live, client-ready deployment.

Delivered outcomes

  • Shipped backend functionality as part of a customer-facing AI receptionist product.
  • Deployed and configured the production website on AWS.
  • Connected product, integration, and infrastructure concerns into one release path.

Need a system like this?

Let’s talk →