AI Engineer

About Kedge Systems Private Limited

Kedge Systems Private Limited is a software consulting company based in Chennai, operating since 2008 with a team of 30 professionals across Chennai and Bangalore, serving European clients.

About the Tempero Project

Tempero focuses on scenario-based validation for simulation testing of autonomous driving systems, encompassing scenario intelligence, causal analysis of scenarios, and software-in-the-loop (SIL) and hardware-in-the-loop (HIL) testing with simulators like CarMaker, DYNA4, and esmini.

Internship Overview

• Role: AI Engineer - Intern
• Openings: 2–3
• Duration: 5–6 months, full-time
• Start Date: December 2026 or January 2027
• Location: On-site at Kedge, SSPDL Alpha City, Navalur, Chennai

Key Responsibilities

• Write production-quality Python code applying AI to autonomous-driving test scenarios.
• Build Python modules and services for Tempero's scenario-analysis and validation workflows.
• Integrate Large Language Models (LLMs) through APIs, including prompt design, structured outputs, tool calling, and retrieval-augmented generation (RAG).
• Develop data pipelines for parsing, cleaning, and analyzing scenario files, simulation logs, and test results.
• Prototype machine-learning and statistical methods, including causal analysis of scenario outcomes.
• Assist in connecting Tempero modules to simulators such as esmini, CarMaker, and DYNA4 in SIL setups.
• Evaluate work by creating benchmark sets, measuring accuracy, and reporting findings.
• Write tests with pytest, document code, and participate in code reviews on Git.
• Demonstrate progress to the team during regular sprint reviews.

Requirements & Tech Stack

• Strong Python fundamentals are prioritized over prior AI experience.
• Solid Python skills including functions, classes, modules, virtual environments, error handling, and reading/writing JSON, XML, and CSV.
• Working knowledge of data structures, algorithms, and object-oriented design.
• Hands-on experience with at least one of NumPy, pandas, scikit-learn, or PyTorch.
• Basic familiarity with LLMs (e.g., calling an LLM API or building a small project with one).
• Comfortable with Git and the Linux command line.
• Clear written and spoken English.
• Ability to present at least one project (GitHub, academic, or personal) in detail.
• Available full-time and on-site in Chennai for at least 5 months.

Qualifications

• Final-year student or 2025/2026 graduate in BE/BTech, ME/MTech, MCA, or MSc in fields such as Computer Science, IT, ECE, EEE, Mechatronics, Automobile, Data Science, or a related discipline.
• Final-semester students require college approval for a full-time internship.

Nice to Have

• Experience building with an LLM framework or SDK (e.g., LangChain, LlamaIndex, or a model provider's SDK), embeddings, or a vector database.
• Experience exposing Python code as a web API using FastAPI or Flask, and packaging with Docker.
• Interest in automotive, ADAS, or autonomous driving, with exposure to OpenSCENARIO, OpenDRIVE, esmini, CARLA, or MATLAB/Simulink.
• Grounding in statistics, probability, or causal inference (e.g., DoWhy).
• Participation in hackathons, Kaggle competitions, or open-source contributions.