Kestra and Smile: Open Data and AI Platforms, Orchestrated

We’re happy to share that Smile is now a Kestra services partner. Its CTO, Jamel Ben Amar, was already writing about Kestra on smile.eu in March 2025, well before the partnership.

Why Smile

Smile has delivered digital projects for more than 30 years. The group has 1,500 people in 9 countries and five practices: AI & Data, Customer eXperience, Business Apps, Cloud Transformation & Management, and Industrial & Product Systems. Alongside Smile, it brings together Neopixl, SensioLabs, UX-Republic, Alter Way, Synotis and creativestyle.

Its AI & Data practice has more than 200 experts and two Data & AI Factories that take 30+ projects a year to production. A large part of that work is what Smile calls digital autonomy: open, independent data stacks that customers control instead of renting from a single vendor. An open source orchestrator built in France fits that brief.

In his article, Jamel compared Kestra with Airflow and found it simpler to configure, easier to read, and built for cloud-native architectures.

“Orchestration has become the core of modern IT and data architectures. By partnering with Kestra, we give our customers an open source, event-driven and declarative solution that automates complex workflows in a few lines of YAML, with full observability.”

Jamel Ben Amar, CTO, Smile Group (translated from French)

What Smile Builds with Kestra

Smile designs and runs data and AI platforms, and Kestra orchestrates what runs on them:

• Open data and AI platforms. Hybrid architectures on open source Modern Data Stack standards, with Kestra self-hosted as the single place to trigger, monitor and rerun every job.
• ETL and ELT pipelines. A typical flow extracts the day’s sales from PostgreSQL, cleans them in Python with Pandas, loads them into BigQuery or Snowflake, notifies the team in Slack or Teams, and reports to Prometheus and Grafana.
• Machine learning. Preparation, training, validation, deployment and monitoring of models in one flow, each step in the team’s own language.
• AI agents. Agents that read from business systems and act under supervised autonomy, run as steps of a Kestra flow with logs, retries and approval steps around them.
• Industrialization. Flows versioned in Git, reviewed and deployed the same way, so each new use case goes from proof of concept to production on the same foundation.

Work with Smile

If you want to build or modernize a data and AI platform with Kestra, Smile’s Data & AI teams can help, in France and across Europe.

• Smile on the Kestra partner directory
• Contact Smile

Consultancies working with Kestra can join the partner program. The first step is getting certified on Kestra Academy.