AI Engineering Intern Generative Models

AI Engineering Intern Generative Models

Company: Katamaran Industries
Job Type: Full-time Internship
Location: Bengaluru, Karnataka Onsite
Duration: 8–10 Months
Experience: 0–2 Years / Freshers
Education: B.E./B.Tech/M.E./M.Tech/M.Sc. in Computer Science, AI/ML, Data Science, Computational Science, Bioinformatics, or a related field

About the Role

Katamaran Industries is looking for an AI Engineering Intern – Generative Models to work on advanced AI/ML systems involving generative modeling, deep learning, scientific computing, and GPU-accelerated workloads.

The intern will work closely with our technical team to develop, train, optimize, and deploy modern generative AI models and supporting ML infrastructure.

This role is particularly suited for candidates who enjoy building models and writing production-quality Python code, rather than working only with pre-built AI APIs.

Key Responsibilities

• Design and implement generative AI architectures for scientific and real-world applications.
• Develop and experiment with Diffusion Models, Flow Matching, Graph Neural Networks (GNNs), Equivariant Neural Networks, and VAEs.
• Build robust and reusable deep-learning training pipelines using PyTorch/JAX/TensorFlow.
• Implement custom loss functions and optimization strategies.
• Explore reinforcement learning and multi-objective optimization techniques for model improvement.
• Develop modular ML tools with clean architecture and reusable components.
• Build data-processing and ML pipelines for large datasets.
• Develop APIs and model-serving components using technologies such as FastAPI.
• Optimize large tensor operations and deep-learning workloads for performance.
• Work with GPU/CUDA acceleration, GPU memory optimization, mixed-precision training, and inference optimization.
• Debug training, data-processing, and model-performance issues.
• Use Linux, Git, Docker, and scripting tools as part of the development workflow.
• Collaborate with researchers and engineers to convert research ideas into working AI systems.

Required Skills

Programming & Engineering

• Strong Python programming skills.
• Good understanding of Object-Oriented Programming (OOP).
• Ability to write clean, modular, and maintainable code.
• Understanding of multiprocessing, parallel processing, or performance optimization.
• Experience with Git/GitHub.
• Comfortable working in Linux environments.

Machine Learning & Deep Learning

• Strong understanding of machine learning and deep-learning fundamentals.

• Hands-on experience with at least one deep-learning framework:

• PyTorch

• JAX

• TensorFlow

• Understanding of neural-network training, loss functions, optimizers, validation, and model evaluation.

Generative AI

Exposure to or hands-on experience with one or more of:

• Diffusion Models
• Flow Matching
• Transformers
• Graph Neural Networks (GNNs)
• Variational Autoencoders (VAEs)
• Equivariant Neural Networks
• Reinforcement Learning
• Custom loss functions

GPU & Computational AI

• Understanding of GPU-accelerated deep learning.
• Exposure to CUDA is an advantage.
• Understanding of GPU memory management and tensor operations.
• Exposure to mixed-precision training or inference optimization is a plus.

ML Engineering

• Experience building ML/AI pipelines.
• Exposure to REST APIs or FastAPI.
• Experience packaging ML models into reusable tools or applications.
• Docker/containerization experience is a plus.

Preferred Experience

Candidates with exposure to the following will be preferred:

• Graph-based scientific data
• Molecular/biological datasets
• Computational biology or scientific computing
• Large-scale ML workloads
• GPU-based model training
• Research projects involving generative AI
• CUDA programming or GPU optimization

Who Should Apply?

This opportunity is suitable for:

• Fresh graduates with strong AI/ML projects.
• Candidates with research or project experience in generative AI.
• Candidates who have implemented models using PyTorch/JAX/TensorFlow.
• Candidates interested in Generative AI, scientific AI, computational biology, and GPU-accelerated machine learning.

Candidates should be comfortable learning advanced AI concepts and implementing them in code.

What You Will Work On

You will get hands-on exposure to:

• Generative AI model development
• Deep-learning research and experimentation
• GPU-accelerated model training
• Large-scale ML pipelines
• Model optimization and inference
• Scientific/structured data
• AI engineering and production ML systems