## We rebuilt the core
Six months ago we sat down and asked a hard question: if we started FlintLab from scratch today, knowing everything we know now, what would we build differently? FlintLab 2.0 is the answer.
This isn't a coat of paint. We rebuilt the execution engine, rewrote the scheduler, and redesigned the dashboard from the ground up. The result is a platform that's faster, more predictable, and we think a lot more pleasant to use every day.
## What's new
### 3x faster pipeline execution
The old execution engine ran each step sequentially unless you manually configured parallelism. The new engine analyzes your pipeline's dependency graph automatically and parallelizes anything that safely can be. In our internal benchmarks, a typical 40-step pipeline that used to take 12 minutes now finishes in under 4.
### Real-time observability
You used to have to wait for a run to finish or refresh a log page to see what was happening. Now every step streams live: logs, resource usage, and status all update in real time on the run dashboard. If something's about to fail, you'll see it happening, not just the aftermath.
### Smarter retries
Flaky steps used to mean flaky pipelines. FlintLab 2.0 introduces adaptive retry logic that distinguishes between transient failures (network blips, rate limits) and real errors (bad config, broken code) so you stop getting paged for problems that would have resolved themselves.
### A redesigned dashboard
Everything from run history to team permissions has been reorganized around the way teams actually use FlintLab day to day: what's running now, what failed recently, and what needs attention. Less digging, more doing.
## Why we built it this way
Performance work is often invisible until it isn't. We kept hearing the same thing from customers: pipelines were reliable, but slow feedback loops were costing teams real time engineers waiting on a run before they could move to the next task. So instead of adding more features on top of the old engine, we treated speed itself as the feature.
## What this means for your team
If you're an existing customer, migration is automatic your existing pipelines will run on the new engine with no config changes required. You should notice faster run times immediately. Some legacy step types will show a "compatibility mode" badge while we finish migrating a handful of niche integrations; we'll follow up with a timeline for those in the next two weeks.
If you're new to FlintLab, this is the best version of the product we've ever shipped, and there's no better time to try it.
## Try it today
FlintLab 2.0 is live now for all customers. Free tier users get the full performance improvements at no cost we wanted speed to be something everyone benefits from, not a paywalled tier.
- **[Try FlintLab 2.0 free →]**