Agentic AI Engineering Intern

Meet the Team

At Ottometric, we're not just keeping pace with the AI revolution in automotive — we're driving it. Our team brings together deep expertise in machine learning, computer vision, and large-scale data engineering, forged at the intersection of two of the most demanding fields in tech: artificial intelligence and autonomous vehicle development.

We specialize in the kind of problems most AI teams never face — managing petabyte-scale, multimodal sensor data, building proprietary KPI computation engines, and deploying AI-assisted root-cause discovery across complex, safety-critical systems. Our platform doesn't just store data; it thinks about it — continuously ranking logs by coverage, novelty, and safety relevance to surface what matters most for model training and validation.

Role Overview

We are looking for a highly motivated AI Coding Assistant Integration Engineering Intern to become our in-house expert on an agentic AI coding platform. You will be the person who builds, tests, and systematically evaluates how the AI coding assistant fits into our engineering workflows — turning a powerful but unfamiliar tool into a well-understood, reliable part of how we build software.

This role sits at the frontier of applied AI engineering. Rather than consuming AI outputs passively, you will design the scaffolding around the AI coding assistant: the persistent instruction files, the MCP server integrations, the CI/CD hooks, the custom slash commands, and the evaluation harnesses that let us measure whether agentic workflows are actually working. You will own this space end-to-end and your findings will directly shape how the wider engineering team adopts the tool.

This role is ideal for a Master’s student who wants to build a genuine portfolio piece in applied AI engineering and who is comfortable operating at the boundary between research and production. 

* While the role is marked as remote, at this time preference will be given to candidates who will join our team on-site in Europe. We will manage all visas necessary.

Key Responsibilities

AI Assistant Workflow Design & Automation

• Design and implement custom AI assistant workflows for recurring engineering tasks: code review, test generation, dependency audits, documentation, and PR triage
• Author and maintain persistent instruction files that give the AI assistant persistent, project-specific context across sessions
• Build and configure custom slash commands to standardise how the team invokes common agentic tasks
• Set up event-driven hooks (PreToolUse, PostToolUse, Stop) to instrument the AI assistant’s actions and connect them to external systems
• Explore and prototype multi-agent orchestration patterns: coordinating parallel sub-agents on complex, multi-file tasks

MCP Server Integration

• Integrate the AI coding assistant with internal and third-party tools via the Model Context Protocol (MCP), enabling it to read design docs, update project management tickets, query databases, and interact with external APIs
• Design the integration architecture: decide what data the AI assistant should have access to, define permission boundaries, and document the integration for the team
• Build a reference integration project that serves as a concrete, reusable example of the AI assistant connected to at least two external services

Evaluation & Benchmarking

• Design and implement an evaluation framework to measure the quality and reliability of the AI assistant’s agentic outputs across different workflow types
• Define meaningful metrics: task completion rate, output correctness, number of human interventions required, context retention across sessions
• Run structured benchmarks comparing AI assistant performance across different instruction file configurations, prompt strategies, and model versions
• Produce clear, reproducible evaluation reports that the team can use to make informed decisions about where to trust and where to constrain agentic automation
• Track regressions or improvements as new coding model versions are released

Documentation & Knowledge Transfer

• Document every workflow, integration, and evaluation methodology in a format that non-specialists on the team can understand and extend
• Produce a “AI Coding Assistant Playbook”: a practical internal guide covering setup, best practices, known failure modes, and recommended configurations for our codebase
• Present findings to the engineering team at the end of the internship with concrete recommendations for permanent adoption