Machine Learning Intern

Enterpret is an AI-native customer feedback platform that turns surveys, reviews, support tickets and community posts into clear product insights. Its customers include Canva, Notion, Descript and Perplexity, and its main engineering team is in Bengaluru. As a Machine Learning Intern, you will build and prototype LLM features with senior engineers and aim to put a real feature into production.

What you'll do

• Build LLM features
• Create RAG pipelines and agents
• Write evals and test plans
• Write production Python
• Debug AI systems
• Ship features to production

What they ask for

• Strong AI/ML project work including coursework, research, open source, hackathons or internships
• At least one project owned end to end
• Clean, testable Python
• Working knowledge of LLMs: prompting, retrieval, agents, fine-tuning or evals
• Systems thinking about data, failures and rollouts
• Available full-time, onsite in Bengaluru

Click on the apply link, fill in your details, add GitHub or project links, upload your resume, and submit.

Tap what's true for you. Your answers are saved for every other job too.

Put these words in your resume and practise these questions before the interview.

Technical

1. How do you implement a RAG pipeline and handle retrieval failures or hallucinations?
2. Can you explain how you would write testable Python code for an LLM evaluation harness?
3. What approaches do you use to debug prompt logic and model output errors?

Scenario

1. What would you do if a newly deployed agent in your pipeline started producing erratic responses in production?
2. How would you approach building a new evaluation metric for an LLM feature when there is no existing ground truth dataset?

HR

1. Tell us about a machine learning project you built completely from scratch and owned end to end.
2. Why do you want to intern at an AI startup like Enterpret rather than a larger tech company?

About Enterpret

1. How does Enterpret process various customer feedback channels like support tickets and reviews to generate actionable product insights?

Practice questions written from the job listing, not this company's real interview questions.

Applied here? Tell the next person what happened. It's anonymous and takes one tap.