About
I am Aaditya Gupta โ a B.Tech graduate in Mathematics & Computing branch at IIT Guwahati and currently working as Data Scientist at Swiggy. I build production-ready ML pipelines and generative-AI solutions. From designing custom CNNs, XGBoost and retrieval-augmented systems to performing A/B Testing, I love turning ambiguous product problems into reproducible, scalable models that deliver measurable business value. ๐๐ถ๐ด๐ต๐น๐ถ๐ด๐ต๐๐: โข Role & focus: Data Scientist @ Swiggy Instamart team. โข Academic foundation: B.Tech (๐ ๐ฎ๐๐ต๐ฒ๐บ๐ฎ๐๐ถ๐ฐ๐ & ๐๐ผ๐บ๐ฝ๐๐๐ถ๐ป๐ด), IIT Guwahati โ strong coursework in Probability & Statistics, Machine Learning, Deep Learning, Linear Algebra, Financial Enginering, Numerical Methods, Algorithms. โข Collaboration & communication: Worked on group projects end-to-end, taken part in ๐ญ๐ฌ+ hackathons and case competitions to solve real-world problems, and contribute actively to the ๐๐ผ๐ป๐๐๐น๐๐ถ๐ป๐ด ๐ฎ๐ป๐ฑ ๐๐ป๐ฎ๐น๐๐๐ถ๐ฐ๐ ๐๐น๐๐ฏ, ๐๐๐ง ๐๐๐๐ฎ๐ต๐ฎ๐๐ถ, where I collaborate on analytics-driven initiatives and knowledge-sharing events. โข Toolkit & practices: Proficient in Python, SQL, Machine Learning, Deep Learning, Data Science, and comfortable implementing models for business outcomes. โข Curiosity & growth: Active focus on advancing in Generative AI, RAG architectures, and deployment-scale ML engineering โ continuously iterating on models, evaluation, and user value. ๐ช๐ต๐ฎ๐ ๐ ๐ฐ๐ฎ๐ป ๐ฑ๐ผ: โข Design and train ML/DL models โข Build end-to-end pipelines: preprocessing, feature engineering, validation, monitoring, deployment (APIs/Streamlit) โข Implement evaluation and model-interpretability routines (ROC AUC, confusion matrices, error analysis) โข Work with imbalanced datasets, missing data handling, and optimization ๐ช๐ต๐ฎ๐ ๐ ๐ต๐ฎ๐๐ฒ ๐ฑ๐ผ๐ป๐ฒ: โข Custom CNN architecture search and hyperparameter tuning โ boosted accuracy from 0.30 โ 0.74. โข Healthcare prediction model โ 0.84 ROC-AUC score, presented insights to stakeholders. โข Toxicity detection pipeline โ achieved 0.96 F1 score using NLP workflows. โข Built spam classifiers, ensemble models, and experimented with A/B Testing โข Experience with model documentation, Sphinx, and GitHub. If you are working on a Gen-AI project, or want to collaborate on research-to-production workflows or have any interesting stuff to discuss related to data science or tech in general โ feel free to reach me.
Work experience
Data Scientist 1 at Swiggy Instamart, Jan 2026 - Present ยท 7 mo, Bengaluru, Karnataka, India
Analyst - Data Science at Sirius AI, Sep 2025 - Dec 2025 ยท 4 mos, Gurugram, Haryana, India
AI Research Intern at Wishtales AI, Mar 2024 - Jun 2024 ยท 4 mos, Bengaluru, Karnataka, India
Advanced Data Analytics Intern at Edunet Foundation, Dec 2023 - Jan 2024 ยท 2 mos, Remote, A 6-week internship in which a healthcare prediction model was created using python. Got to learn a lot through regular sessions by industry experts and created a presentation report summarising the whole work.
