Killian Bourgeat

Callenviais an AI voice assistant for **businesses**:

it answers **calls**, qualifies each request, and transforms

missed calls into opportunities. Designed for

restaurants, hotels, salons, and independent services.

The AIanswers calls in natural language, manages

reservations, answers frequently asked questions, and switches

toSMS/WhatsApp/Instagramto extend the conversation.

Everything is controlled from a centralized **dashboard**.

My role covers the entire product: design, architecture,

development, deployment.

Real-time LLM architecture: speech recognition (STT),

response generation (LLM), speech synthesis (TTS), latency

management for natural phone interaction. Telecom integration

(incoming calls, transfer, fallback to human team).

Unified conversational agents across calls, SMS, WhatsApp,

and Instagram Direct. Customer context maintained across

channels, automatic request qualification, reservation booking,

FAQ management.

Per-tenant configuration: custom instructions, hours, FAQs,

tone of voice, language. The AI adapts to each brand and

type of business (restaurants, hotels, beauty, services).

Multi-tenant professional portal: call tracking, complete

transcripts, automatic summaries, reservation management,

statistics, assistant configuration.

Multi-tenant, data isolation per establishment, secure

REST API, horizontal scaling, monitoring, telecom and

messaging integrations (WhatsApp Business API, Instagram

Messaging API).

PostgreSQL, OpenAI, Anthropic, STT/TTS models,

Twilio (telephony), WhatsApp API, Instagram API,

Stripe, Docker, AWS, WebSocket.

oneSiteis a SaaS I designed and developed from scratch:

anyone can create a complete e-commerce website in

minutes, simply bychattingwith an AI.

The user describes their business in a few sentences.

The AIgenerates a complete e-commerce site in real time,

hostedandreadytosell. The user subscribes,

then manages their catalog, orders, shipments, and

statistics from an AI-powered back-office.

An end-to-end project: from product UX to AWS infrastructure,

covering the entire lifecycle of an AISaaS: design,

development, monetization, operation.

Full-stack LLM architecture: prompt analysis, multi-step

content generation, automatic page formatting, live preview

synchronized by WebSocket during the conversational session.

OpenAI / Anthropic integration, advanced prompt engineering,

API cost management.

LLM-powered back-office: sales data aggregation,

anomaly and opportunity detection, contextualized product

recommendations, visitor behavior analysis. Scoring and

recommendation generation pipelines.

Per-tenant data isolation, secure REST API, role management,

horizontal scalability.

Page, block, media, style, and theme editing by the end

user. Complete back-office: catalog, orders, inventory,

shipments, sales statistics.

EKS (Kubernetes) deployment, automated CI/CD, horizontal

scaling, monitoring, automated domain and SSL certificate

management per tenant. Asset storage (S3), CDN.

PostgreSQL, MongoDB, OpenAI, Anthropic, Stripe, Docker,

Kubernetes (EKS), AWS (S3, EC2), CI/CD.

Long-term mission for the completerefactoringof critical applications in the aeronautics sector: modernization of the technical foundation, version migrations, distributed architecture, and real-time event orchestration.

Complete rewrite of existing applications: Java migration, Spring Boot, Angular version upgrade to Angular 21. In-depth refactoring, removal of technical debt, alignment with modern cloud-native standards. Steering technical choices and mentoring the team on breaking changes.

Design and development of high-load distributed services as part of the refactoring. Resilience management, observability (structured logs, metrics, distributed tracing). Code reviews, definition of team technical standards.

Implementation and operation of Kafka infrastructures: topic design, partitioning strategies, optimized producers/consumers for throughput, error handling, dead-letter queues, event replay. Real-time event processing on high-volume streams.

Complete overhaul of business interfaces: complex components, state management with RxJS, performance optimization, REST integration. Migration and modernization of the front-end foundation, tooling, test coverage.

Deployment on Azure (AKS), Kubernetes orchestration, CI/CD pipelines, Docker containerization. Demanding production environment with availability and performance constraints.