Apr 2025 – May 2025 · 1 mo · Dubai, United Arab Emirates · On-site

• Awarded Pre Placement Offer

Dec 2023 – Feb 2024 · 2 mos · Remote

• 1. Researched & led the development of AI analytics dashboard, indirectly processing data from 50K+ users
• 2. Designed front-end for analytics dashboard, incorporating 15+ key metrics and 5+ graphical representation
• 3. Collaborated with founders on wireframe design, conceptualizing multiple features and KPIs for dashboard

Aug 2022 – Nov 2022 · 3 mos · Remote

• 1. Mentored a cohort of around 20 students about different aspects of python programming, Machine learning and AI.
• 2. Took sessions on specific topics such as KNN and python programming and Insights about ML
• 3. The cohort rated me 4.6/5 and Considered me resourceful and knowledgeable.

Dec 2021 – Oct 2022 · 10 mos

• Worked on making Zomato a better place through:
• Computer Vision (OCR) for handwritten Images:
• Worked on optical character recognition by using the following tools and libraries:
• TR-OCR, Tesseract, keras_ocr, attention_ocr, and various other tools such as trdg for synthetic data generation.
• Data Science (Fraud detection):
• Worked on fraud detection (Model development + Deployment) at the user level by following tools:
• Feature Selection and Extraction, Model Selection, MLflow with Fast-API Model development, Vegeta Load Testing, Butterfly Setup, Monitoring on Grafana
• NLP (Embedding clustering):
• Clustered Embeddings using various available libraries including sentence BERT, Word2Vec, Para2Vec, etc, used Density-based Scanning for clustering and t-SNE for projecting embeddings to two-dimensional data for visualization of clusters.

Jun 2021 – Dec 2021 · 6 mos

• Introduced several improvements and experiments hence making the product a grand success by analytics of the product
• Reduced under-delivery of clicks by calculating contributions from several factors such as CTR, Shutdown, and Sell-able Score and hence implemented a solution that decreased UD considerably
• Inventory management for ads by creating tableau inventory dashboard with excellent visual represented data to effectively understand the situation
• Created several tableau and Redash dashboards using Presto-SQL for analyzing the data in the best possible way and taking appropriate actions

Jan 2021 – Jun 2021 · 5 mos

• Created Tableau dashboards for effective inventory management. Created ETL pipelines for making various derived tables as per the requirements. Helped with various data extraction and summarization with the help of tools such as Excel and SQL.

Aug 2020 – Sep 2020 · 1 mo · Work from home

• Developed a machine learning model to predict the fantasy points of players in the squad for the next match
• Binary Integer Programming method was applied for the selection of 200 fantasy teams having a high probability of winning
• The model for team selection was deployed and provided very good results when it was tested on 10 live matches

Jul 2020 – Sep 2020 · 2 mos · Gurugram, Haryana, India

• Implemented a semi-supervised learning model for object detection using the Faster-RCNN model with Resnet-50
• The idea of the noisy student model was applied to achieve higher mAP when labeled/annotated data is small in size
• The mean average precision score of the model improved by about 10% over the corresponding supervised method
• The project was fast-tracked to implementation and received a stellar Letter of Recommendation from the company

Jul 2020 – Aug 2020 · 1 mo · Work from home

• Scraped, cleaned the data, and applied various text augmentation techniques to increase the size of the data
• Experimented with a different state of the art architectures including seq2seq and pre-trained BERT and GPT-2 models
• The best model involved the reinforcement learning technique and was deployed and had a rapid response time

Jun 2020 – Aug 2020 · 2 mos · India

• Predicting Stock Price for the next day using LSTMs (Kaggle) Closing prices form NSE Data
• Implemented a deep learning model using LSTMs to predict closing prices for the next day using data from the previous 60 days
• The model was able to predict prices with very good accuracy. The R-squared value on test data was around 0.96
• News Article Classification using CNNs (Kaggle) IBN news articles classified into 5 categories
• Implemented a 1-D convolutional neural network architecture with Glove word embeddings to classify news articles
• The model was very fast and was able to classify the articles with an accuracy of 92% over the given dataset

Dec 2019 – Jan 2020 · 1 mo · Sydney, New South Wales, Australia

• Implemented an end to end machine learning model for predicting biological age from a MEG dataset of size 1.4 TB (Cam-CAN)
• Deployed a pipeline, including a sequence of Pre-Processing, Representation, Projection, Vectorization, and Regression
• A comparative analysis of 4 different pipelines was done, and the best pipeline provided state of the art predictions
• Work was praised and appreciated by the professor and received an honorarium award for the work done during the internship

Sep 2019 – Aug 2020 · 11 mos · Greater Delhi Area

• Elected unanimously and negotiated overall increment of 249.1% in the maintenance budget with many novel initiatives
• Implemented an online system to keep a track and resolve any maintenance related issue of the hostel.

May 2019 – Aug 2019 · 3 mos · Greater Delhi Area

• Novel Bimetallic catalyst for O.D.E.C. reaction
• Prepared a novel bimetallic catalyst for efficient and cost effective production of ethylene by oxidizing ethane using CO2
• Analyzed the performance of the catalyst on a bench scale reactor at different temperatures using gas chromatography
• Thermodynamic analysis of the reaction and characterization of the catalyst is being done to study behavior of the catalyst
• Selectivity of the catalyst was found to be 10% higher than the precious bimetallic catalysts used for the reaction currently

Nov 2018 – May 2019 · 6 mos · South Delhi, Delhi, India

• Established organic waste composting hubs in 15 societies, 20 schools across NCR generating revenue of over 15 lakhs
• Led a team of 10+ executives to organize various mega-events and sessions under our campaign named Adopt a park
• Efforts appreciated by Cheif Minister of Delhi and Swachh Bharat Abhiyaan and featured in TOI and Doordarshan

Jul 2018 – Oct 2018 · 3 mos · Greater Delhi Area

• Approached 50+ startup companies, food joints and educational institutes; designed customized events for the sponsors
• Successfully converted deals to procure total funding of INR 1.25 lakhs to ensure smooth conduct of the fest