Job Description

What this Job Entails

This is a hands-on internship within the automation engineering function, supporting Astreya's product engineering teams. The role is focused on building AI agents and system integrations that automate operational workflows.

The intern in this seat contributes production code under mentorship. Work is scoped, reviewed and delivered in the same manner as that of permanent engineers, with support proportionate to experience. The emphasis is on automation outcomes — reducing manual effort in operational processes — rather than on general development support.

Scope

Delivers one or more scoped automation deliverables per phase of the internship, end to end, under mentor review

Works within an established codebase and existing engineering standards

Escalates blockers promptly; independent judgment is expected within the boundaries of the assigned scope

Your Roles and Responsibilities

AI agent development

• Design, build and refine AI agents that automate operational workflows — contextual enrichment, routing and triage decisions, knowledge retrieval and guided resolution steps
• Develop prompts and agent logic that produce reliable, structured, machine-consumable output, and handle failure conditions predictably
• Implement tool-calling and multi-step agent workflows where an agent must retrieve data or execute an action rather than return a response
• Build evaluation methods for agent behaviour — labelled test sets, accuracy measurement and confidence threshold tuning — so that agent performance is demonstrated rather than asserted
• Ensure agents surface supporting reasoning with every recommendation and record the outcome of each human decision, in line with platform governance requirements
• Model and framework choices are made per use case; the requirement is to build agents competently against whichever API the team has selected, not familiarity with any single provider

Integration and workflow automation

• Build and maintain integrations between the platform and enterprise systems, including ITSM, HR, observability and configuration management sources
• Implement authentication, pagination, rate limiting, retry and reconciliation logic to production standards
• Define and document field mappings between systems, and maintain integration tests that detect upstream changes before they affect downstream consumers
• Automate manual operational processes identified by the team, and measure the effort removed

Delivery practice

• Maintain clear commit history, pull request documentation and runbooks for all delivered work
• Present completed work to the team at the end of each phase and articulate the design decisions taken
• Participate constructively in code review, both as author and reviewer
• Other duties as required. This list is not a comprehensive inventory of all responsibilities assigned to this position

Required Qualifications/Skills

• Completed a B.Tech, BE , MCA or BSC in any discipline
• Strong working knowledge of Python — functions, classes, modules, package and environment management, HTTP clients, JSON handling and structured error handling; able to read and extend an existing codebase
• Demonstrated experience building with large language model APIs, provider-agnostic — prompting, structured output and ideally tool calling or multi-step agent flows. Academic, personal or hackathon projects are acceptable evidence
• Sound AI/ML fundamentals — embeddings, vector search and retrieval-augmented generation, the distinction between prompting and fine-tuning, model limitations and hallucination, and the principles of model evaluation
• Working knowledge of REST APIs — authentication, pagination, rate limiting and error semantics — and the ability to work from API documentation independently
• Version control with Git, including branching and pull request workflow
• Basic SQL
• Clear written communication; candidates will be asked to explain a technical decision in writing as part of the assessment
• Experience building an API service using FastAPI, Flask or equivalent
• Familiarity with containers and Docker
• Exposure to Google Cloud Platform or an equivalent public cloud
• Experience with an agent orchestration framework such as LangChain or LangGraph
• Working knowledge of PostgreSQL and data handling with pandas
• Exposure to enterprise ITSM platforms or to IT operations processes
• A public code repository demonstrating relevant work