Athenian was one of the best-known engineering-metrics products. It measured DORA metrics for teams, not for individuals. Today the product is gone. Its website, athenian.com, now redirects to the company's GitHub organisation. The app login does the same. There was no shutdown announcement. The founder, Eiso Kant, moved on to build Poolside, an AI company.
Many teams that used Athenian still need the same thing: clear delivery metrics without developer surveillance. This post covers three questions. What happened to Athenian? What changed in DORA since its site went quiet? And what should replace it — a new dashboard, or something that closes the loop?
What happened to Athenian
The facts are short. Athenian raised $6M in March 2022, led by Point Nine. Its stated position was strict: team-level metrics only, never individual metrics. Public engineering activity on its GitHub organisation stopped in mid-2023. By late 2025, athenian.com redirected to GitHub. The company published no shutdown notice. Review listings on G2 and Gartner still show it as an active product. It is not.
Athenian is not alone. The standalone metrics category is consolidating hard. Pluralsight Flow has a published end-of-life date of December 31, 2027, and closed renewals in June 2026. Code Climate split its business in late 2024 and narrowed Velocity to enterprise deals. The pattern is consistent: a metrics dashboard on its own struggles as a company. Measurement keeps its value only when it is attached to something that can act on the numbers.
DORA moved on while the dashboards stood still
Athenian's homepage froze in the four-metrics era: deployment frequency, lead time for changes, change failure rate, and time to restore. That era has ended.
The DORA research program changed its model in 2025. The report is now titled the State of AI-assisted Software Development. It tracks five delivery metrics — the four keys plus rework rate — and it retired the famous elite, high, medium, and low tiers. Teams now map to seven performance archetypes instead. The old benchmark tables that most metrics tools still ship are one full generation behind the research.
The 2025 report also settled the AI question with data. Ninety percent of surveyed developers now use AI at work, with a median of two hours per day. The report's central finding: AI is an amplifier. It raises throughput, and it magnifies instability wherever the underlying delivery system is weak. The 2024 report saw the early version of the same effect — AI adoption correlated with a 1.5% drop in throughput and a 7.2% drop in delivery stability. Faster code generation does not fix a slow delivery system. It stresses it.
Why the replacement should not be another dashboard
A metrics product can only observe. It shows you a lead-time chart. It cannot merge the stuck pull request, staff the review bottleneck, or roll back the failed deployment. Athenian's own homepage promised "autonomous improvement." A read-only analytics layer cannot deliver that promise. Nothing that only watches can.
This is the lesson of the consolidation wave. Gartner expects software engineering intelligence platforms to reach 50% of engineering organisations by 2027, up from 5% in 2024 — but the buyers are choosing platforms, not point dashboards. The measurement has to live where the work happens: next to the delivery pipelines, the team capacity, and, increasingly, the AI agents doing part of the work.
That is how we built DORA metrics on the Scrums.com platform. The metrics sit inside the same plane that orchestrates delivery: the same platform runs developer productivity analytics, capacity planning, and the AI agent harness. When a number moves in the wrong direction, the lever to correct it is on the same screen — add capacity, change the working agreement, or adjust what the agents are allowed to ship. Measurement and action in one system.
What to keep from Athenian's philosophy
Athenian got one big thing right, and it is worth keeping: measure teams, not individuals. Individual-level metrics create surveillance, and surveillance corrupts the data. Engineers optimise the number instead of the outcome. DORA's own research has always operated at the team and organisation level for the same reason.
Our DORA implementation holds that line. Deployment frequency, lead time, change failure rate, failed-deployment recovery time, and rework rate are computed per team and per deployment stream. No individual leaderboards. If a tool offers you a per-developer stack ranking, it is measuring the wrong unit.
A practical migration checklist
If you are replacing Athenian, or any retired metrics tool, check five things before you commit:
- Current DORA model. The tool must report the five 2025 metrics, including rework rate, not the 2021 four-keys table.
- Team-level measurement. No individual metrics. This protects both your data quality and your engineers.
- Deploy-event accuracy. Metrics from Git activity alone miss the deployment boundary. The tool must ingest real deploy events.
- AI-impact visibility. With 90% of developers using AI daily, you need to see AI-assisted throughput and its effect on stability, not just raw volume.
- A path from insight to action. Ask the vendor what happens after the chart turns red. If the answer ends at an alert, you are buying half a product.
The metrics category's history is now clear enough to act on. The dashboards that only watched are shutting down. The research they reported against has moved. Pick the measurement that lives inside your delivery system — and can do something about what it finds.
See how DORA metrics work on the Scrums.com platform, or talk to us about your delivery setup.
