
Remote, India, Nagpur, India
This is not a typical internship, and we are not looking for someone to make slides or clean spreadsheets.
Glitter Technology Ventures Pvt. Ltd. (www.glitterlabs.com) is offering a Research Internship for undergraduate and graduate students who are serious about AI research, working across Edge AI, Vertical AI, and large-scale content generation.
We want someone who can read a paper in the morning, reproduce it by afternoon, and explain what it gets right, and what it misses, by the end of the week.
If you think critically, experiment rigorously, build what you study, and want to leave with research you can point to, keep reading.
About the Role
This is a structured 3–6 month internship. You’ll own a real research project, work weekly with a senior researcher, and collaborate with our engineering team and founders. You’ll leave with a publication-track result or research artifact.
You are:
Owning an assigned research project aligned with your interests and our roadmap.
Running experiments to test ideas, reproduce published work, and evaluate new approaches.
Building software, including prototypes, pipelines, and tools that turn ideas into working systems.
Producing a research output: a paper draft, technical report, benchmark, or open-source release.
Presenting your work to the founders and team at the end of your tenure.
We move fast. Deadlines are real commitments.
Location
Remote, India
Nagpur, MS, India (local candidates preferred)
We work in person, remote, or hybrid for speed, clarity, and tight feedback loops.
Duration
3–6 months, with a minimum commitment of 3 months.
What You Get
Assigned research project: aligned with your interests and our roadmap.
Weekly mentorship: 1:1 sessions with a senior researcher.
Stipend: scaled to duration and contribution level.
Compute: full access to lab compute and research infrastructure.
Research presentation: present your work to the founders and team.
Publication-track outcome: a paper draft, report, benchmark, or open-source release.
Certificate and recommendation: based on the work you deliver.
Path to full-time: exceptional interns may be considered for a full-time role.
Core Responsibilities
Research & Experimentation
Scope research questions with your mentor into clear milestones.
Review prior work and reproduce key results.
Design and run experiments, ablations, and evaluations on lab compute.
Analyze results honestly, including negative ones.
Technical Analysis
Evaluate models and approaches on accuracy, latency, cost, and deployability, especially for edge and domain-specific use.
Build and use benchmarks to measure quality, factuality, and robustness.
Validate findings with the engineering team against real-world constraints.
Software Development
Write clean, modular, tested code for prototypes, experiments, and evaluation pipelines.
Build tools that speed up data processing, training, and benchmarking.
Work with engineering to turn promising prototypes into production-quality components, including edge and on-device targets.
Use Git, code reviews, documentation, and reproducible environments.
Profile and optimize for performance, memory, and compute efficiency.
Research Writing
Document your work continuously, not just at the end.
Produce a final research artifact and present it to the team.
Turn findings into clear reports, articles, and technical explainers.
Execution Discipline
Own your research plan and make progress between check-ins.
Keep code, experiment logs, and decisions organized and reproducible.
Flag blockers and slips early.
Skills & Requirements
Required:
ML foundations: a solid grasp of machine learning and deep learning fundamentals.
Python: strong skills plus hands-on experience with PyTorch, JAX, or a similar framework.
Software engineering: the ability to write maintainable, tested code and use Git.
LLM familiarity: an understanding of transformers, fine-tuning, RAG, quantization, or evaluation.
Paper fluency: the ability to explain a paper’s contribution, method, and limitations.
Clear writing: the ability to communicate complex results simply.
Nice to have:
Prior research: a thesis, lab work, published papers, or competitions.
Infrastructure: experience with GPUs, cloud platforms, containers, or experiment tracking.
Open-source: contributions, a GitHub portfolio, or technical blog posts.
AI-assisted workflows: experience using AI tools to accelerate research and development.
What matters most: curiosity, rigor, engineering ability, learning velocity, and follow-through.
Qualifications
Currently pursuing, or recently completed, a Bachelor’s or Master’s in Computer Science, Engineering, IT, Mathematics, Statistics, Data Science, or a related field.
Strong projects, code, or writing count for more than grades.
Why Join Us
Do real research and build real systems, not toy exercises.
High ownership from day one, with direct access to the founders.
Combine research, engineering, and writing in one role.
Leave with work you can publish, present, and build on.
Most students study AI. Few get to research it and build what they discover.
Interview Process
Before you apply, read how our interview process works: https://www.glitterlabs.com/interview.html
Note:
During the interview, please answer questions using the STAR (Situation, Task, Action, Result) format. Be specific, concise, and quantify impact wherever possible.
Our interview page includes a guide and a worked example: https://www.glitterlabs.com/interview.html#star
We read these responses carefully and use them to understand how you think, execute, and create meaningful outcomes.