Background Agents for Engineering Teams
Delegate work to agents that start on their own. Your models, your infrastructure, no markup on tokens.
Trusted by engineering teams
How Felan Works
Developer attention is the bottleneck. Felan takes the work that shouldn't need anyone watching.
Starts on its own
From an event, a schedule, or a person on the team.
- Event5xx alert
- ScheduleMon 09:00
- PersonSlack ask
03:12Session started
Investigate 5xx spike on checkout-api
Works unattended
On shared infrastructure, not a laptop, so no one keeps a session open.
Your laptop, closedFelan’s infrastructureRunningInvestigate 5xx spike on checkout-api
Claude Opus 5.5
- Environment ready03:12
- Read logs and the last 12 deploys03:14
- Traced the errors to deploy #143203:19
- Connection pool runs out under load03:23
- Testing a fix03:26
Comes back for review
As a PR, a fix, or a report — at the review point you choose.
+12 −3Raise connection pool limit
checkout-api #1433
- Checks passed
- Regression test added
- Review requested from on-call
Felan in #incidents · 03:31
Traced the 5xx spike to deploy #1432. The fix is ready in #1433.
What Teams Delegate First
Start with one repo and one repetitive task. Delegate more as trust grows.
Connected Once, for the Whole Team
Agents take triggers from the tools you already use and report back in the same place.
GitHub
GitLab
Bitbucket
Azure DevOps
Linear
Jira
Asana
ClickUp
Notion
Confluence
Slack
Microsoft Teams
Email
Figma
Vercel
Supabase
PostHog
Testiny
Any MCP server
Built for Unattended Work
Unattended agents need an always-on, isolated place to run. Felan runs it for you.
Always on
Agents pick up work at 3 a.m., and the whole team sees what they did.
- 03:12Alert investigated, draft PR #1433 opened
- 04:40Flaky auth tests fixed
- 06:00Dependencies updated, tests green
Isolated by default
Each run gets a clean environment with only the access it needs.
checkout-api session
- RepositoryRead and write
- CI logsRead
- Developer laptops
- Production keys
Many at once
Run agents in parallel, each from the same known setup.
Running in parallel
Managed cloud or your own infrastructure
Start on Felan's cloud, or run Felan on your own servers.
Felan cloud
Managed for you
Same agent
Your infrastructure
Self-hosted
Same agent
Predictable Token Costs
No markup, any model, and a cost on every task.
See what every task costs
Every session records its model and its cost.
| Session | Model | Cost |
|---|---|---|
| Investigate 5xx spike on checkout-api | Claude Opus 5.5 | $1.84 |
| Fix failing CI on main | Claude Sonnet 5 | $0.62 |
| First-pass review of PR #1432 | GPT-5.5 | $0.31 |
| Replace deprecated date helpers | Kimi K3 | $0.27 |
| Flaky test sweep | Gemini 3.5 Flash | $0.14 |
| Weekly dependency updates | DeepSeek V4.1 Flash | $0.09 |
| This week | $3.27 |
Your models, your choice
All major providers, plus local and private models.
Default model
Search models
- Claude Sonnet 5Anthropic
- GPT-5.5OpenAI
- Gemini 3.5 FlashGoogle
- DeepSeek V4.1 FlashDeepSeek
- Kimi K3Moonshot AI
- Private modelYour endpoint
No markup on tokens
0%
Bring your own keys or subscription, or use Felan credits billed at provider rates.
An agent built to use fewer tokens
41.9%
Lower aggregate model cost for Felan Code across seven controlled extension benchmarks. Read the study
Teams Already Delegating QA
Bugzy, Felan’s QA agent, tests every change for these teams without anyone watching.
The tool became so independent — it picks up the Jira tickets ready for testing on its own and just reports the results in Slack. Every morning I see that the main test cases have passed, and I feel comfortable knowing the system is tested constantly.
Maya Serafimova, QA Lead, GoodShape
1,000+
Automated test runs in 5 months
< 12 min
Average test cycle
97%
Execution success rate
We ship features daily with a 4-person team because Bugzy catches regressions before we even review the PR. It's like having a QA engineer that never sleeps.
Milko Slavov, Founder & CEO, Bugzy AI
100%
PRs verified automatically
< 30 min
From deploy to QA results
0
Manual test maintenance
One Price for the Whole Team
Every account starts with a 14-day Team trial. Model usage is always billed separately.
Individual
For trying Felan on your own repo.
€0
- 1 member and 1 repository
- 200 agent minutes per month
- Bring your own keys or subscription
Team
For the whole team. One price, not per seat.
€99per month, per team
- Unlimited members and repositories
- 5,000 agent minutes per month
- Up to 5 concurrent workers
- Felan credits at provider rates
Enterprise
For security reviews and larger rollouts.
Custom
- Self-hosting on your infrastructure
- SSO and SAML
- Custom limits and dedicated support
Common Questions
Something else on your mind? Ask the team.
What is Felan?
Felan runs background AI agents for software engineering teams. Agents start from an event, a schedule, or a person, work without anyone watching, and run on shared infrastructure — Felan's managed cloud or your own — instead of developers' laptops. A person on the team reviews the result.
Does Felan replace my coding agent?
No — keep using it for the hard problems. Felan takes the work that shouldn't need anyone watching.
What should we delegate first?
Start with one repo and one repetitive task with a result worth checking, such as failed build triage, dependency updates, first-pass code review, or incident investigation. As the team builds confidence, delegate more and move the review point later.
Is it safe to let agents run unattended?
Each run happens in its own isolated environment with limited access, not on a developer's machine next to their files and credentials. Every session is visible to the team, and results come back for a person to review.
Which models can we use?
All major model providers, including local and private models. Use a cheaper model for simple work and a top model for hard problems, and switch when prices or quality change.
How does Felan keep token costs low?
There is no markup on tokens: bring your own keys or subscription, or buy Felan credits billed at provider rates. Felan Code, the open-source agent at the core, is tuned to use fewer tokens without lowering quality, and you can see what every agent and task costs.
Can we run Felan on our own infrastructure?
Yes. Use Felan's managed cloud, or run Felan on your own infrastructure with the Enterprise plan.
Does Felan replace engineers?
No. Engineers own direction, judgment, and review. Felan takes the work that shouldn't need anyone watching, so they can keep their attention on the hard problems.
How much does Felan cost?
Individual is €0 for one developer and one repository. Team is €99 per month for the whole team, not per seat. Enterprise has custom terms. Every new account starts with a 14-day Team trial, and model usage is always billed separately — by your provider, or through Felan credits at provider rates.
Keep Your Attention for the Hard Problems
Start with one repo and one recurring task.
