Generative AI/ML Internship

Digital Blaize

Applied 902
Application Deadline 10 days left
Eligibility
Engineering Students
Undergraduate
Postgraduate
Management
Arts, Commerce, Sciences & Others

Refer & Win

MacBook, iPhone, Apple Watch,
Cash and more!

Refer now Know more

Recruitment Process

Screening

Interview

You'll have a live interview with our hiring team, analyzing your responses in real-time to provide evaluators with deeper insights. This helps us have a more objective conversation about your experience and fit. Please Complete the Google form https://docs.google.com/forms/d/e/1FAIpQLSeIghGjzGZCv3aqQYgUI46gcdwofdLeXj-tYFvOYXVYf2JTvA/viewform?usp=header

Interview

Interview round

Details

Responsibilities:

  • Fine-tune open-source LLMs (LLaMA 3, Mistral, GPT-2, Quen2 ) using LoRA/QLoRA techniques for domain-specific use cases
  • Collect, clean, preprocess, and structure training datasets for model development
  • Design and run prompt engineering and evaluation workflows to optimize model outputs
  • Build and test RAG (Retrieval-Augmented Generation) pipelines using vector databases like ChromaDB or FAISS
  • Experiment with RLHF (Reinforcement Learning from Human Feedback) and model optimization strategies
  • Develop evaluation benchmarks measuring reasoning, accuracy, and generation quality
  • Assist in deploying fine-tuned models via APIs (FastAPI/Flask) for integration with web products
  • Collaborate with the engineering team, document experiments, and present findings regularly

Requirements:

Must-Have

  • Pursuing or recently completed a degree in Computer Science, AI/ML, Data Science, or a related field
  • Strong Python programming skills (NumPy, pandas, scikit-learn)
  • Hands-on familiarity with Hugging Face Transformers, PyTorch, or TensorFlow
  • Understanding of ML fundamentals: training loops, loss functions, evaluation metrics
  • Basic knowledge of LLMs, embeddings, and fine-tuning concepts

Good to Have

  • Experience working with LLaMA, Mistral, Falcon, or similar open-source models
  • Familiarity with LoRA, QLoRA, PEFT parameter-efficient fine-tuning methods
  • Prior exposure to LangChain, LlamaIndex, or vector databases
  • Basic understanding of REST APIs or Flask/FastAPI backends
  • Access to a GPU environment (Colab Pro, Kaggle, or local GPU) is a plus

What You'll Gain:

  • Real-world experience building production-grade LLM pipelines
  • Hands-on training under an experienced AI team at Digital Blaize
  • Certificate of Completion + Letter of Recommendation
  • Potential Pre-Placement Offer (PPO) for high performers
  • Exposure to the full AI lifecycle — from data collection to model deployment
If an employer asks you to pay any kind of fee, please notify us immediately. unstop does not charge any fee from the applicants and we do not allow other companies also to do so.

Important dates & deadlines?

  • 4 Jul'26, 12:00 AM IST Registration Deadline

Additional Information

Stipend

Work Detail

Working Days: 5 Days

Schedule: Flexible Work Hours, Monday to Friday,

Job Type/Timing

Job Type: Work From Home

Job Timing: Full Time

Perks

Job Offer

Certificate of Completion

Letter of Recommendation

Flexible Hours

Hybrid Working

Counselling Support

Pre-Placement Offer

Voice your opinion by leaving a feedback & your rating

*This opportunity has been listed by Digital Blaize . Unstop is not liable for any content mentioned in this opportunity or the process followed by the organizers for this opportunity. However, please raise a complaint if you want unstop to look into the matter.
Raise a Complaint