2 months ago
Bengaluru, IndiaMid Level
Responsibilities
- Build agentic workflows for support intake, classification, context gathering, reproduction, diagnosis, response drafting, remediation, verification, and follow-up.
- Develop tool-using agents that inspect runs, logs, authentication state, configuration, and provider behavior and execute bounded recovery actions safely.
- Create evaluation suites for diagnostic correctness, resolution quality, safe escalation, customer communication, and end-to-end task completion.
- Build human-in-the-loop systems with explicit ownership, uncertainty, approvals, handoffs, and failure states.
- Develop observability, memory, and feedback loops that improve agents using real support cases without repeating mistakes.
- Debug, reproduce, isolate, mitigate, ship, communicate, and verify production issues directly with customers.
- Turn recurring customer problems into workflows, tools, product capabilities, and durable fixes.
- Partner with Support, FDE, Product, and Engineering to move prototypes into reliable production workflows with guardrails and human fallback.
Requirements
- 2–4 years of professional software engineering experience.
- Experience building production systems with language models, tool calling, retrieval, structured outputs, multi-step workflows, or agent memory.
- Ability to use evaluations, traces, failure analysis, and rapid iteration to make AI systems dependable.
- Experience as a backend, platform, or integration engineer who has shipped and operated production software.
- Ability to debug across SDKs, HTTP, OAuth, webhooks, queues, data stores, logs, and third-party APIs.
- Ability to turn ambiguous failures into minimal reproductions, root causes, and durable fixes.
- Interest in working directly with technical customers and translating vague problems into software systems.
- Ability to model support functions using states, tools, permissions, handoffs, verification steps, and escalation rules.
- Ability to distinguish what should be automated from what should remain human while increasing autonomy safely.
- TypeScript or Python production experience is optional.
- Experience with AI agents, copilots, workflow automation, evaluation infrastructure, developer platforms, API products, support engineering, SRE, or incident response is optional.
- Familiarity with OAuth 2.0, webhooks, rate limits, Postgres, queues, cloud observability, Plain, Zendesk, Intercom, or Salesforce is optional.
- Public technical writing, open-source contributions, or strong debugging notes are optional.
Benefits
- Full-time role based in Bengaluru, working US timings.
Tech Stack
Categories
About Composio
Composio builds a developer platform that lets AI agents connect to and take actions in third-party apps such as GitHub, Gmail, Notion, and Salesforce. It offers APIs and SDKs that handle authentication, context, execution, and reliability so teams can embed tool use into agentic workflows. Founded in 2023 and headquartered in San Francisco, the company operates as a privately held SaaS vendor focused on enterprise and productivity integrations.