Forward Deployed Engineering Internship (Fall 2026 & Winter 2027)
Required skills
- on-prem lifecycle
- helm charts
- upgrades
- release train
- rollback paths
- customer forks
- temporal workflows
- k8s init ordering
- storage
- auth-refresh pipelines
Nice to have
- Temporal
- Postgres
- cloud secret managers
- self-hosted delivery
- kots
- airgap
Role summary
Composio is seeking a Forward Deployed Engineering Intern for Fall 2026 & Winter 2027 to help define reliable, repeatable on-prem agent deployment for their AI infrastructure. The intern will work with the on-prem lifecycle, including Helm charts and releases, debug across the stack (Temporal, Kubernetes, storage, auth), and support security efforts like CVE triage and patching. Ideal candidates are curious about Kubernetes and Helm, have experience building in uncontrolled environments, are AI-native with opinions on agents, and are comfortable in Linux. Bonus points for experience with Temporal, Postgres, cloud secret managers, or self-hosted delivery.
At Composio, we are building infrastructure that allows agents to communicate with the tools you use for work including Github, Gmail, Notion, Salesforce, etc. We are a lean team of engineers wrangling problems from context to search, that help us provide the most capable bridge between your agents and your tools.
We raised a $25M Series A from Lightspeed with some incredible angels like Guillermo Rauch (CEO of Vercel), Dharmesh Shah (CTO of Hubspot), Gokul Rajaram. Beginning of this year we 3x our ARR with customers like Google, AWS, Mercor, Mintlify, Glean and Emergent.
running stateful agent infrastructure inside someone else's cluster is new territory. as our FDE intern, you'll help define what reliable, repeatable on-prem agent deployment looks like across customers running us in their own environments.
What you'll do?
get hands-on with the on-prem lifecycle: helm charts, upgrades, the weekly release train, rollback paths.
help pull diverged customer forks back onto one supported chart, with guardrails.
contribute to onboarding so a new customer deploys in days, not weeks.
debug across the stack: temporal workflows, k8s init ordering, storage, auth-refresh pipelines.
support the security surface: CVE triage, patching, RCAs.
"Must haves"
if you are very good, nothing is a must per-se
you've built in environments you don't fully control — homelab, VPC, self-hosted — and like figuring out why systems break.
curious about kubernetes and helm; you want to go deep.
ai native — you've built with language models and have opinions on agents and MCP
comfortable in linux; you can write a clear RCA.
human — you build trust and admit what you don't know.
BONUS: temporal, Postgres, cloud secret managers, self-hosted delivery (kots/airgap), open source or hackathon wins.
