PATIL AKSHAY VILAS VAISHALI

AI-powered comprehensive tax automation platform serving 5,000+ businesses

• Architected AI-powered tax classification engine using ensemble machine learning models, achieving 98% accuracy and reducing manual intervention by 65%
• Built intelligent GST reconciliation system processing 1M+ transactions monthly, reducing processing time from 5 days to 2 hours
• Developed NLP-driven ITR data extraction pipeline that automated 90% of tax document processing with 96% accuracy
• Implemented deep learning algorithms for Capital Gains computation with 95% precision, improving compliance accuracy by 40%

NEURAL NETWORK FOR CLASSIFICATION OF STARS AND GALAXIES

• Created a pipeline to download FITS files from Sloan Digital Sky Survey(SDSS) sky observatory and converted files into images (numpy array)
• Developed a star-galaxy classification framework that uses a Neural Network model on these images
• Trained model using Convolution Neural Network with the help of TensorFlow library

Google Cloud Platform implementation for text detection and recognition

• I started this project as a way to learn GCP and it has since grown into a fully end-to-end python project
• I have learned many skills after this and been I’m very proud of having this in my profile

Various machine learning and computer vision projects

• Developed machine learning based python program to detect mobile manufacturer names using mobile images
• Applied statistical methods for image generation in multiple NFT projects
• Created neural network project for currency note classification
• Applied Deep Learning methods for Character Recognition in Devanagari Languages
• Galaxy Detection using Artificial Intelligence with radio telescope data

Leading cross-functional engineering teams in developing enterprise SaaS platform for comprehensive tax automation

• Led cross-functional engineering teams serving 5,000+ businesses with comprehensive tax automation encompassing GST, ITR, and Capital Gains processing
• Architected and deployed AI-powered tax classification engine using ensemble machine learning models, achieving 98% accuracy and reducing manual intervention by 65%
• Spearheaded development of intelligent GST reconciliation system processing 1M+ transactions monthly, eliminating compliance discrepancies
• Built NLP-driven ITR data extraction pipeline that automated 90% of tax document processing with 96% accuracy across multiple tax schedules
• Managed team of 12 engineers and data scientists, driving adoption of MLOps practices that achieved 99.8% system up-time
• Established automated model retraining pipelines using TensorFlow and PyTorch, maintaining system accuracy above 95%
• Mentored 8 junior engineers in advanced AI/ML techniques, leading to 3 internal promotions
• Leveraged cloud-native architecture on AWS to reduce infrastructure costs by 30% while supporting 15,000+ concurrent users

Led end-to-end data science initiatives across multiple high-impact projects

• Spearheaded cross-functional collaborations with engineering, product, and business teams to translate complex data insights into actionable business strategies
• Architected and deployed ML-powered invoice processing system that reduced manual processing time by 85% and improved accuracy to 99.2%
• Developed enterprise-scale Python solution for TDS Reconciliation processing 10M+ transactions monthly with 100% compliance
• Built NLP-powered data extraction platform processing 50,000+ tax documents annually with 95% accuracy
• Implemented scalable Flask API services enabling seamless integration with downstream applications, reducing data processing costs by 40%
• Mentored 4 junior data scientists in advanced ML techniques and production deployment best practices

Applied statistical methods for image generation in multiple NFT projects

• Developed custom algorithms for procedural image generation using statistical methods
• Worked on multiple NFT projects implementing creative AI solutions

Artificial Intelligence research and development for document processing automation

• Applied and evaluated OCR problems using deep learning algorithms such as CNN, FRCNN
• Designed, developed and tested NLP algorithms such as RNN, LSTM to extract insights from text
• Used TensorFlow Serving to deploy models for production use on AWS/GCP cloud services
• Developed document information extraction automation for Bank Account Statements, CIBIL Reports, KYC Documents
• Built end-to-end machine learning pipeline for Electricity Meter Reading and data masking on sensitive information

Research in computer vision and astronomical data analysis

• Contributed to research strategy within the AI department
• Identified methods for Text Detection and Recognition in printed documents
• Worked on Galaxy Detection using Artificial Intelligence with radio telescope data
• Generated and preprocessed image data from radio sources for Deep Learning classification

Major in Computer Science with Specialization in Machine Intelligence

University Department of Computer Science, one of the oldest departments with a Master’s level programme in Computer Science. Focused on advanced machine learning, deep learning, and artificial intelligence techniques.

The University of Mumbai is one of the oldest and premier Universities of India

Studied at Bhavan’s College, one of the oldest and top colleges in Mumbai. Strong foundation in mathematics, statistics, and analytical thinking that forms the basis of my data science career.

Advanced certification in applying machine learning techniques to astronomical data

Credential ID: UC-dfa4bf16-906b-48db-8e11-a2b806c8680f

Leadership training and community service initiatives

• Completed Leadership Training Program organized by University of Mumbai
• Organized various social activities including Blood Donation Camps, Cleanliness Drives, AIDS/HIV Awareness Campaigns
• Assisted blind and physically handicapped students with examination support