Hi HN! I'm Antoine, co-founder and CEO at Okay (YC W20). We are building analytics for engineering teams and we have released our new application on Github Marketplace.
Getting good metrics on engineering work is hard. On the one hand, you can build your own pipeline by ingesting API data from Github, JIRA, your CI tool, etc. Then you have to ETL that data, link it to teams and services, and finally select a dashboarding tool to display metrics. That gives you ultimate flexibility at a high cost of maintenance. On the other hand, you can go with out-of-the-box products but a lot of them focus on very specific metrics, such as DORA Metrics, PR throughput, etc. - not necessarily what you'll be interested in tracking, and you might even disagree with the metrics definitions or ethos of the product.
We've gone through several iterations of Okay and we think we now have the right balance between flexibility and ease-of-use. We solve ingestion + ETL and we give dashboarding features, but we leave users free to define metrics using our no-code query builder (with plain SQL coming soon). We specifically avoid forming opinions on what users should actually be building in terms of dashboards - in that sense, we are closer to an observability tool than to "development analytics" products. We are seeing a range of use-cases, such as incident tracking (with JIRA and PagerDuty data), sprint reviews (JIRA / Github), velocity & DORA (mainly Github + deployment data). We are building on BigQuery, DBT for the ETL and our custom ingestion code for API syncs. Our frontend is using Vue with a mix of D3.js and common charting libraries.
We'd love to get the community feedback on this current product approach! This Github app makes it easy to start on our free plan.
