I'm a Computer Engineering student in Mumbai working on information retrieval, long-term memory for language models, and evaluation methodology — the question of how you measure whether a system actually does what its designers claim.
My work spans machine learning, backend systems, and constraint optimisation: designing retrieval architectures and training classifiers, building async services over PostgreSQL, and modelling scheduling problems with CP-SAT solvers. I write the documentation and the evaluation as carefully as the code — architecture references, threat models, cost models — because that is how I find out whether I actually understand a system.
I try to make my work falsifiable. My projects tend to ship with a written account of where they fall short: which components were defective, which were never properly exercised, and which carried the result.
I'm preparing for graduate study in machine learning, and I'm looking for research and startup internships in retrieval, memory, or applied ML systems.
Python · PyTorch · scikit-learn · FastAPI · PostgreSQL · OR-Tools · Java · Docker · Linux
18deepnar@gmail.com