About Aviva

We are a Mexico City-based technology startup following a unique approach to unsecured credit for underserved communities. Aviva’s physical onboarding kiosks build a bridge to those living in remote communities and provide a documentless 5-minute video onboarding for everyone in need of liquidity. This asset-light acquisition strategy, combined with a superior risk model based on computer vision and natural language processing, will allow Aviva to offer the lowest interest rates available to today’s underserved.

What you will do

We already have a decent automation: we are using Pants to have reproducible builds of the diverse services contained in our monorepo, DVC for reproducible experiments and avoid regressions, and a GitHub actions pipeline for automatically deploying our services to Kubernetes (managed through TrueFoundry).

However, given Aviva's fast growth, we feel the need to go one step above: improve the observability of our applications, optimize our CI/CD pipelines, implement comprehensive testing strategies for our services, automating training runs, ... So we are looking for someone who would help us do all that, and make the team run smoother.

The ML engineers will continue to be responsible for testing, deploying and monitoring the models they create.

Your role would be to:

• Improve self-service tools, or create new ones.
• Introduce best practices to the team.
• Jump-in when the team runs into technical problems out of the ordinary.
• Coordinate with data engineers (for instance, in the deployment of a feature store).
• (Optional) Take on some ML problems of your own, to understand the workflow and difficulties of the ML engineers.

This will enable us to deploy with more confidence, and to focus more of our time in the improvement of our ML algorithms.