How you'd work
Small pods, direct access, no account layer.
You'll work in a dedicated pod of 3–6 people spanning design, engineering, and ML — embedded with the client, not rotating through a bench. You'll talk directly to the client's engineers and stakeholders, not through an account manager relaying messages. Most people here have shipped more than one production system since joining.
Who you talk to
You work directly with the engineers and designers on your project — not a separate account manager relaying messages.
Team structure
Dedicated pods, sized to the project. We don't staff by the bench — the same people stay on from kickoff to production.
Project length
Engagements range from a focused 6–8 week build to multi-year platform partnerships — several clients have stayed with us for years.
Requirements
What we're actually looking for.
About Codemonk
Codemonk is a fast-growing product engineering and IT services firm that partners with clients to build scalable digital solutions. From startups to enterprises, we help businesses solve real-world challenges through cutting-edge technology and deep domain expertise. We've partnered with brands like Unilever, Kawada, Tata Power, greytHR, DMG, Experien, and more.
About The Role
We are looking for a motivated Machine Learning Intern to join our team and assist in building and optimizing generative AI models. This internship offers hands-on experience in developing AI applications with the latest machine learning and natural language processing technologies.
Key Responsibilities
A) Generative AI
- Algorithm Design & Development: Design, implement, and debug machine learning algorithms with a focus on transformer-based models and generative AI applications
- LLM API Integration: Build applications leveraging LLM APIs from OpenAI, Anthropic Claude, Gemini, and other foundation model providers
- RAG Pipeline Development: Construct and fine-tune Retrieval-Augmented Generation (RAG) systems using vector databases such as FAISS, Pinecone, etc.
- AI Application Deployment: Learn to deploy AI applications and models into production environments with proper monitoring and optimization
B) Deep Learning & Computer Vision
- Model Training & Tuning: Working knowledge of training and fine-tuning deep learning models, including adapting pretrained models to new tasks.
- Experimentation: Run and track experiments to compare models, parameters, and datasets, and draw clear conclusions from the results.
- Data Preparation & Audit: Help prepare, clean, and audit datasets for relevance, quality, and alignment with the problem being solved.
- Model Evaluation: Evaluate models using appropriate metrics and identify where and why they underperform.
- Optimization & Deployment: Basic exposure to making models efficient and helping move them into production.
Required Technical Skills
- Programming Proficiency: Strong foundation in Python programming with knowledge of any generative AI frameworks, including LangChain, LlamaIndex, and Hugging Face libraries, is a plus.
- LLM Integration Experience: Familiarity with LLM APIs (OpenAI GPT-4, Anthropic Claude, Cohere) and interest in building AI applications
- Fundamental Architecture Knowledge: Understanding of Deep Learning Fundamentals, including transformer-based models, attention mechanisms, and their optimization for various AI workloads
- Embedding & Vector Systems: Basic understanding of embedding models (OpenAI, sentence-transformers) and vector databases
- Analytical Excellence: Strong problem-solving, debugging, and analytical skills with proven ability to work with complex algorithms and datasets
Compensation & benefits
Benefits that come with the role.
Compensation is discussed and finalised directly with you at the offer stage, based on experience and scope.
Learning & development
An annual budget for courses, conferences, or certifications — used at your discretion, not pre-approved case by case.
Flexible working hours
Core collaboration hours with the client's team; the rest of your day is yours to structure.
Health coverage
Comprehensive medical insurance for you and your dependents from day one.
Ownership from day one
You're on the pod that ships the system — not a bench waiting for staffing.
Meet the team
Some of the team you'd be joining.
“You're not maintaining someone else's roadmap here — you own the system, from architecture to the 2am page. That's what makes the engineering here different from most services work.”
Hiring process
Four steps, no surprises.
Telephonic Screening
A 20–30 minute call with our talent team about your background and the role — not a script, a conversation.
Assignment
A scoped take-home, typically 3–4 hours, built around a real class of problem we work on — designed to see how you think, not to extract free work.
Technical Rounds
Two rounds with engineers you'd actually work with — one on system design, one on your assignment and past work.
Cultural Round
A conversation with leadership about how you work and what you're looking for next.
Most candidates hear back within 2 weeks of their final interview.
Ready to build this?
Applications go straight to our team — no recruiter queue in between.