aritro shome — other

← back to main

you cannot be possibly that interested in me.

achievements

  1. role of ai in material science research — co-authored paper on AI-based statistical techniques for material innovation. 1st place, Brahma Prakash Memorial Material Science Paper Presentation 2023, IGCAR Kalpakkam
  2. ipl 2025 predictor — random forests, mlps, monte carlo. won BrainDead 2025, IIEST Shibpur
  3. abler — real-time accessibility assistant fusing object detection, depth, speech and gesture. winner, SXC eXabyte Hackathon (demo)
  4. MentL — mental wellness assistant. best design, Hacknoware 2025
  5. PPG encoder & SMS spam classifier frontends — awarded, Tensor Quest, Instruo 2025

smaller builds

  1. alponagen — procedural generator for alpona (bengali folk art) via fractal layering; 13k-image dataset on kaggle
  2. ml & nlp from scratch — regression, knn, k-means, mlp, bow, tf-idf, cbow/skipgram, bpe, naive bayes, lstm sentiment — pure numpy/pytorch, no black boxes
  3. thomas raw — agentic travel planner (gemini + langchain); plans, budgets, identifies locations from photos
  4. vidvaan — research paper aggregator across DBLP / ArXiv / OpenLibrary / OpenAlex with gemini summaries
  5. where am i? — web-based geography guessing game
  6. RNN music generation, engagement prediction, image-filter experiments — github

certifications

  1. deep learning specialization
  2. convolutional neural networks
  3. sequence models
  4. structuring ml projects
  5. improving deep neural networks
  6. genai with gemini
  7. vertex ai studio

skills

languages: python, java, c/c++, javascript, html/css

ml / ai: pytorch, scikit-learn, numpy, pandas, langchain, nltk

tools: git, flask, matplotlib, firebase, google cloud

misc & extra curricular

[+] last updated: june 08, 2026