My name is Alexis Guyot. I design and lead complex cloud-native platforms dedicated to research, data science and artificial intelligence environments, placing automation and user experience at the heart of every architecture.
A systems engineer by training, I work at the intersection of technical architecture, infrastructure industrialization and operational IT strategy. My approach is guided by a simple conviction: build robust, readable and automated systems so their users can focus on what matters, without being held back by underlying complexity.
At Groupe GENES, I led the technical evolution of the Onyxia platform — a multi-tenant Kubernetes infrastructure in production serving scientific entities (ENSAE, ENSAI, CREST, CEPE). This platform pools shared resources (GPU, object storage, open-source services) and allows users to provision self-service environments within a secure, automated and observable framework.
Automation is at the core of how I work. I design platforms built to last, evolve and remain manageable over time. Building infrastructure that makes complexity as readable as a YAML file sums up my approach to systems pretty well.
Beyond the technical layer, I place great importance on cross-team collaboration, process clarification and user adoption — because infrastructure only has value when it is understood and mastered. I am currently pursuing the PMP certification to structure and deepen my project leadership skills according to international standards.
Built after the AWS Cloud Practitioner cert. Hosted on S3, served globally via CloudFront with OAC enforcing CDN routing. Lambda triggers on each page load to increment a DynamoDB visitor counter. CORS restricts the Lambda URL to the production domain.
Full CI/CD via CodePipeline + GitHub. Domain via Route 53, SSL/TLS via Certificate Manager.

Calculates estimated travel time for a given distance across all human transport modes, compared to the speed of light. Frontend on S3, backend via Lambda + API Gateway, calculations in Python/boto3, results stored in DynamoDB.
→ Live Demo
Map UI to add GPS coordinates. Authentication via Cognito. REST API built with Lambda + API Gateway stores coordinates in DynamoDB. Continuous deployment via Amplify.

Speedlight fully deployable via Terraform : Lambda + DynamoDB, S3 static hosting with for_each + etag=filemd5, API Gateway REST avec CORS, and a Python script patching the API URL into the HTML via Terraform output.
Personal Kubernetes cluster built on Raspberry Pi 5 nodes, designed to explore cloud-native technologies and make a catalog of on-demand deployable services available. GitOps with ArgoCD, SSO via Keycloak, secrets management via Vault, Onyxia service catalog.
The lab also explores KubeVirt for VM workloads on Kubernetes, OpenTelemetry for distributed tracing, and Rancher for cluster management.