Back to careers

Local runtimes, workflows and reliability

Agent Systems Engineer

Build the runtime and execution systems that let agents complete useful work with clear permissions, durable state and dependable recovery.

We hire according to current project needs. We welcome expressions of interest in these roles; timing and scope depend on our priorities.

Where you’d apply it today

Vowdo

Bounded local business work in execute and improve modes — scoped tools, inspectable execution, improvement only from verified feedback.

What you’ll do

  • Design the agent runtime around explicit interfaces for model calls, tools, sessions and execution environments; support long tasks across restarts and context resets.
  • Build permissioned filesystem, process, browser and API tools. Enforce filesystem and network isolation, protect secrets and require review for consequential actions.
  • Make state transitions durable and writes idempotent. Handle queues, concurrent tasks, cancellation, timeouts and retries without duplicating effects.
  • Make each run inspectable through traces and audit trails; reproduce failures and evaluate recovery, tool correctness and prompt-injection resistance in controlled environments.
  • Keep execution separate from improvement: version tools and instructions, evaluate verified feedback, and review changes before they affect future runs.
  • Measure reliability, latency and cost; choose scheduling, caching and model-routing strategies that keep the system practical to operate.

What you’ll bring

  • Strong backend or systems programming in Python or TypeScript, with Linux, process, filesystem and networking fundamentals.
  • Experience with durable workflows, queues or distributed systems; reason clearly about concurrency, retries, idempotency and crash recovery.
  • Practical agent integration: model APIs, tool protocols, context limits and session state. Separate probabilistic model decisions from enforceable runtime rules.
  • Security judgment: least-privilege access, filesystem and network isolation, secrets handling, authorization and untrusted tool output.
  • Ability to diagnose failures from traces, build controlled integration tests and measure whether the system completes work correctly.

Helpful experience

  • MCP clients or servers, authenticated service connectors, OAuth and scoped credentials; browser or computer-use automation.
  • Containers, OS sandboxes or microVMs; experience comparing isolation, resource limits and startup overhead.
  • Local application packaging and updates, OpenTelemetry, fault injection or multi-agent scheduling with explicit ownership and bounded concurrency.

Why you’ll find it rewarding

  • Shape the execution layer: how agents use tools, retain progress and recover when work goes wrong.
  • Turn security and reliability constraints into a system people can confidently use for real work.
  • Build reusable infrastructure for today’s workflows and the products that come next.

How to apply

Send a short introduction and links to relevant code, projects or a portfolio; a CV is optional. Tell us what you built, a failure you investigated, and how you checked whether it worked. We value demonstrated ability over particular degrees or years of experience.

Apply by emailOr write to us directly: contact@dpintelli.com

Opens your email app with a prepared draft addressed to us.