Internship · Remote (India)

AI Research Intern

Cehpoint is hiring for this role: internship, Remote (India), ₹10,000–₹20,000/month. Explore RAG, evaluation and agent reliability.

About the role

Conduct rigorous benchmark evaluations, hallucination audits, prompt optimization experiments, and retrieval evaluations to maximize production model reliability.

Investigate retrieval, evaluation and reliability for our AI systems. You'll run experiments, measure quality, and help us make our AI more accurate and trustworthy.

What you should know

  • Foundations of machine learning and NLP: attention mechanisms, transformer models, embeddings
  • LLM evaluation metrics and methodologies (RAG triads, Ragas, faithfulness, answer relevancy)
  • Advanced RAG strategies: parent-child chunking, reciprocal rank fusion, hybrid BM25 + dense search
  • Python scientific stack: PyTorch, NumPy, Pandas, Scikit-learn

What you should be able to do

  • Building automated benchmark harnesses and reproducible evaluation pipelines
  • Statistical analysis of model failure modes across adversarial inputs
  • Synthetic data generation and fine-tuning dataset preparation
  • Authoring clear, concise technical experiment summaries and actionable recommendations

How you work

  • Relentless intellectual skepticism toward LLM outputs and dedication to eliminating hallucinations
  • Eagerness to track and evaluate state-of-the-art model releases and open weights
  • Curiosity regarding agent planning algorithms and multi-agent consensus protocols

Skills we check

  • Analytical mindset
  • Python
  • Interest in ML/LLMs
  • Writing up findings

How we select for this role

We read every application ourselves. Here is what happens after you apply.

  1. Apply with a short form

    Paste a link to your CV and answer five questions in your own words. No sign-in and no file upload. About fifteen minutes.

  2. Integrity check

    When you press Check my answers, an automated check confirms the answers were typed by you. You see the result straight away, before you submit.

  3. Choose your project date

    If no cheating is detected you are shortlisted, and we email you a link to choose the day your project starts.

  4. 5-day practical project

    On that day our team contacts you with the brief: real work, 5 working days, reviewed by an engineer. For this role: Run benchmark evaluations on 3 RAG chunking/retrieval strategies and present quantifiable analysis.

  5. Offer

    If your work solves the problem well, you receive your offer letter by email and in the portal.