I’m currently working at a startup and planning to build a data closed-loop system using Flyte. Our pipeline will cover the entire lifecycle: data collection, cleaning, labeling, feature embedding, model training, and offline inference. Since we are starting from scratch with no legacy constraints, would you recommend going straight with Flyte 2.0 or sticking with Flyte 1.x? Also, do you have any specific advice for this setup?

My take would be to use Flyte2. Thats what the team is moving forward with. But in the mean time try out flyte 1 in your pre-prod/development so you know the use cases and gather some familiarity with the tools. Flyte 2 OSS will be available to general public in the first quarter of 2026. Also they have Union 2 managed service which you can try, which already uses Flyte 2.Reach out to @powerful-horse-58724 if you need access to the managed service offering

Flyte 2.0 on union is generally available now - OSs version is soon. Check out flyte 2 docs here https://www.union.ai/docs/v2/byoc/user-guide/ Lots of examples https://github.com/flyteorg/flyte-sdk/tree/main/examples

Thank you for the great reply and tag Debasis! Like Ketan and Debasis mentioned you can try Flyte 2.0 on a union cluster right now, and it should be OSS within Q1. I think everyone who starts using doesn't want to go back to Flyte 1. but totally get still getting started with v1 since it has current OSS. If you're up for a meeting, we'd love to discuss which options would fit best for you. https://www.union.ai/meeting