Model Ops

Enabling data scientists to build better models and deploy faster
ML Platform
Developer Tool
Duration
6 months
Scope
Product strategy, UX design
Role
Sole designer
What was the problem?

With the API-only approach, it is very hard for customer data scientists to learn and use the machine learning tool provided by C3. The goal of this project is to build a GUI tool for machine learning APIs to increase efficiency and ease of use, so that we can increase customer conversion rate and reduce total hours of customer support.

Current state: API-only

What were my responsibilities?

As the sole designer on the team, I covered the end to end process from primary research, lofi exploration, concept validation to hifi design.

SOLUTION
Track experiments

ModelOps automatically track parameters, metrics, datasets, and other artifacts of experiments.

Visualized model comparison

Helps user quickly compare experiments and runs and identify fine-tune strategies.

Review models and track model lineage

Stakeholders and collaborators can easily understand key details about a given model.

Easy deployment

Deploy the model and manage deployment status with one click.

If you are interested in learning more about this project, please view full case study here:

View full case study