Romain GÉRARD

Head of Engineering Hands-on - 12 people, Qovery, remote

Engineering organization

• Re-organizing teams to avoid knowledge silos and reduce management work due to product priority changes
• Increased written communication to give autonomy without communication overhead
• Normalized teams' delivery process, cadence and output format to align delivery with company objectives
• Put in place a FinOps practice & manage budgets for tools, services, and infrastructure Costory

Technical direction

• Define and carry the engineering strategic roadmap, AI adoption, and an agentic software factory inside teams
• Aligning the technical roadmap between sales, customer support, product and compliance stakeholders

Cross-functional alignment

• Work closely with Product, Sales, Customer Success, Marketing, and the CEO to ensure business objectives are met
• Involved with customers to POC product initiatives
• Managing conflicts between divergent incentives

Culture

• Created an engineering café for all to share and learn beyond the organizational structure
• Ensuring operational excellence by formalizing post-mortems and making them systematic
• Contributed to setting expectations for the product engineer role

Staying operational - Hands-on

• Staying operational and compensating for an organizational void when needed
  • Help deliver features to close deals and POC with clients
  • Ramp up on new tech - Kubernetes Gateway, AI
  • Supporting teams by sharing maintenance work and tackling tech debt

Lead Staff Backend Engineer, Qovery, remote

Development:

• Rewrote and designed the multi-service backend from scratch to meet the reliability, security, and scalability required to run a business SaaS Kotlin, Spring Boot, PostgreSQL, Redis, REST API, JWT, Prometheus

• Created agents/operators running on clients' Kubernetes clusters to extract real-time information (logs, metrics, status, ...) Rust, Tokio, gRPC, Kubernetes

• Created network gateways to command and control clients' cluster agents to provide users with on-demand real-time interactions Rust, Tokio, WebSocket, gRPC, Remote shell, ECDSA

• Worked on the Qovery engine to deploy and manage the infrastructure of our clients Rust, Helm, Terraform, Kubernetes, AWS, GCP, Scaleway

• Design and implementation of new product features Kotlin, Rust, Terraform, Helm, AWS, GCP, Scaleway

• Contributed to open-source projects by fixing upstream issues

  • https://github.com/hyperium/h2/issues/633
  • https://github.com/rust-lang/futures-rs/pull/2686
  • https://github.com/kube-rs/kube/pull/1112
  • https://github.com/nats-io/nats.rs/pull/177
  • https://gitlab.com/glts/viaspf/-/merge_requests/3

Infrastructure:

• Allowed the team to deploy multiple times a day by putting in place CI/CD with automatic testing GitLab CI, Kubernetes, Docker

• Set up observability to ensure correct behavior of the platform and fast incident resolution Prometheus, Grafana, Loki, Alloy, Kubernetes

• Put in place security (authentication + encryption) on our message bus to guarantee client data safety Auth0, JWT, NATS, TLS, Let's Encrypt, Kubernetes

• Put in place CI to support ARM64 and AMD64 builds of all our components/delivery to allow clients to reduce their cost by using the ARM64 CPU architecture GitLab CI, BuildKit

Lead:

• Mentored newcomer developers to ramp up on the infrastructure, technical stack and product

• Worked with product to help define & prioritize new features and challenge technical difficulties

• Participated in the recruitment process, managing underperformance

• Participated in layoff decisions ... due to the previous economic situation

Lead Software Resiliency Engineer, Criteo in Paris

• Responsible for the NoSQL perimeter (>= 4000 server assets) Cassandra, Memcached, Couchbase

• Grew the team from 2 to 6 people, created a roadmap and built adhesion for quarterly, 9-month, and 12-month plans (interviews)

• Improved low-latency monitoring of Memcached C++, eBPF

• Stabilized and ensured resiliency of the Kubernetes platform

• Participated to put in place a new S3 object storage service with Ceph and migrated users to it Rook Ceph

• Participated to put in place security for Elasticsearch, Kubernetes, etcd LDAP, JWT, OIDC, TLS

• Developed a Memcached proxy and POC'd in production a replacement for the legacy Couchbase platform ScyllaDB

Senior Software Resiliency Engineer, Criteo in Paris

• Developed software to automate auto-failover of critical services Python, Couchbase

• Dropped commercial support (2M€ license fees) for Couchbase with new resiliency gains

• Took ownership of Elasticsearch and created a new stack for it Go, Java, Elasticsearch

• Put in place Kubernetes at Criteo from scratch and migrated 3000 machines to it to reduce toil induced by managing bare-metal machines Chef, flannel

• Worked to create a plan to migrate the S3 object store (pithos + Cassandra) to a more robust solution Minio, Ceph

• Handling the opening of 2 new datacenters and closing of 2 legacy ones without disruption of services

Software Resiliency Engineer, Criteo in Paris

• Managed NoSQL infrastructure (more than 4000 bare-metal servers) to provide it as a service Memcached, Couchbase, Cassandra

• Integrated a new monitoring system and discoverability to solve resiliency issues Prometheus, Consul

• Developed new monitoring probes to ensure the SLA and provide monitoring as a service by default Java, patch in Prometheus

Software Developer, Criteo in Paris

• Developed a time series database backend over Cassandra for Graphite Python, Java, Cassandra, Graphite

• Managed infrastructure as code Chef

• Wrote a patch merged upstream in Cassandra to meet Criteo R&D performance needs Java, CASSANDRA-13418

Software Developer, Orange in Sophia Antipolis

• Developed and put in production a machine learning model to detect anomalies in time series Java, Python, Spark, Cassandra, Hadoop, MapReduce, Jenkins

• Developed a Web API and UI to offer this ML model's anomaly detection as a service HTML/CSS, JavaScript, Bootstrap, D3JS, Java, Tomcat

Trainee Software Developer, Amadeus in Sophia Antipolis

• Study and integration of a NoSQL database to improve fault tolerance within a critical distributed system C++, Couchbase
• Development of tools to improve code quality Clang, LLVM, Static analyzer

Trainee Software Developer, Arkoon Security Network in Lyon

• Network probe to detect an application’s traffic C++, Python, MongoDB, HTML5/CSS3, CoffeeScript

Trainee Software Developer, Corail Software in Bidart

• Rich web application to manage the flow of mail for La Poste with a Bayesian network C#, Silverlight