Role Overview

We are seeking a highly motivated AI/ML Intern with a focus on Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG).  You will be working at the frontier of generative AI, helping us explore, build and optimize systems that bridge the gap between static models and dynamic, private data.

This role is ideal for a student looking to apply theoretical knowledge to a production-grade environment and build a significant portfolio piece.

Key Responsibilities

Architect & Optimize RAG Pipelines: Improve retrieval accuracy using vector databases (e.g., Pinecone, Milvus, or Weaviate).

LLM Implementation: Experiment with and fine-tune prompts and workflows using models like GPT-4, Claude, or Llama 3.

Data Engineering: Assist in the cleaning, chunking, and embedding of proprietary datasets.

Evaluation: Implement benchmarking frameworks to measure hallucinations, faithfulness, and relevancy of model outputs.

Qualifications

Academic Standing:  Graduate student (Masters/PhD) in Computer Science, Data Science, or a related technical field.

GPA: 3.5 or higher preferred.

Technical Skills: Proficiency in Python and experience with AI frameworks (e.g., LangChain, LlamaIndex, PyTorch, or TensorFlow).

Domain Knowledge: A solid understanding of transformer architectures and the mechanics of RAG.

Soft Skills: A research-oriented mindset with the ability to troubleshoot complex, non-deterministic systems.

What You Will Gain

Mentorship: Direct access to several AI chief architects and weekly 1-on-1 growth sessions.

Portfolio Impact: Significant contribution to a live AI project that you can showcase to future employers.

Flexibility: We respect your academic schedule and offer flexible working hours.

Future Opportunities: Top performers will be prioritized for future full-time, paid openings. 

We are a diverse group of AI/CV and big data experts, three professors, and several PhDs, looking to add an intern to our team to develop cutting-edge computer vision software.

Company Description

Ottometric is revolutionizing the validation and training of Advanced Driver Assistance Systems (ADAS) with our AI-driven solution. Our integrated software platform streamlines and automates the entire process, addressing big data management, training, and real-world validation challenges. By doing so, Ottometric helps companies save millions in development costs, enhance reliability, meet safety standards, and accelerate time to market.

Role Description

This is a full-time on-site role based in Novi Sad, Serbia, lasting six months. You will help design and implement advanced algorithms for a variety of computer vision tasks, both based on traditional geometry-based methods and using cutting-edge 3D reconstructions models.

The Ideal Candidate

You are a hands-on computer vision engineer who thrives on solving inherently hard 3D reconstruction problems. You understand modern Transformer/VLM architectures and how to apply them in efficient ways, but you also understand why traditional CV methods are sometimes required. You have worked with real-world sensor data (camera, lidar, IMU) that needed wrangling, because real-world data is never clean.

Basic Qualifications

• Current student in Computer Science, AI, or related graduate degree programs.

• Proven record of solving real-world 3D CV problems, such as structure from motion, camera pose estimation, and SLAM

• Expert software development skills, with industry experience in building production-quality software

• Excellent written and spoken communication skills in English

• Ability to work collaboratively in a fast-paced environment

⠀

We will cover all relocation and visa requirements.