aaparikh - Personal Portfolio

Email
aaparikh.me@gmail.com
Github
aaparikh
LinkedIn
aaparikh
Medium
@aaparikh_
Google Scholar
Atharva Parikh

About Me

Innovating with AI & Data Science

I am an AI & ML Developer, Product Strategist, and Co-Founder passionate about building AI-driven solutions that create real-world impact. My expertise spans AI research, LLM development, cloud computing, and multi-agent systems.

Key Highlights:

🚀 Co-Founder & CPO at BrainGnosis – Developing LLM-powered multi-agent AI systems.
💰 Managing $600K+ in AI contracts with leading tech and research organizations.
🏥 Building AI-powered healthcare assistants at Purdue University for early disease prediction.
🤖 Created CookGPT – an advanced RAG system reducing search times by 85%.

I am always open to collaborations that push the boundaries of innovation. Let's build the future together!

Download Resume

Currently learning and practicing

Cloud Computing

Breaking down the complexities to build scalable and secure applications.
AI/ML Development

Building intelligent systems that learn from data and improve over time preserving privacy. Main focus on healthcare AI.
Data Enginnering

Building data pipelines, data lakes, and data warehouses to enable data-driven decision making.
Data Analysis

Analyzing data to extract insights and make data-driven decisions.

Technical Skills

Experience

BrainGnosis

AI and Product Strategy (Co-Founder, CPO)

West Lafayette, IN, USA Jan 2025 – Present

Created an LLM Agent framework - SciBorg to rapidly deploy MultiAgent systems for organizations, with features like Modularity, advanced benchmarking, Finite State Automata (FSA), and dynamic memory for Agent to Agent communication.
Managing $600K+ worth of AI contracts with leading tech and research companies.
Spearheading the development of an AI operating system for enhanced process efficiency in enterprises.

Purdue University

AI and Software Developer

West Lafayette, IN, USA Jan 2024 – Present

Developed AI-powered healthcare assistants for Cleveland Clinic using LangChain and time-series models (LSTM,TimeVAE, Online Learning) to predict brain hemorrhage risks with 97% accuracy, enabling early intervention.
Built advanced RAG automation pipeline for personalized medical insights and automated alert systems.
Built pluggable AI tools, such as automated PubChem integration for chemical information access and instrument drivers for controlling and running chemical instruments in AI-driven smart labs.
Co-authoring research papers on AI applications in healthcare and scientific research, in collaboration with the NIH.

Cook Medical

Machine Learning Engineer

Bloomington, IN, USA Jun 2024 – Aug 2024

Engineered a "Complaints Forecasting" model using Azure ML pipelines, achieving an R² score of 0.92 to accurately predict priority products for quality checks and hence minimize the losses.
Developed "CookGPT" leveraging Azure AI (Search Service, Datalake, OpenAI API) to implement an advanced search and retrieval (RAG) system within Cook Medical's internal knowledge base, drastically reducing the search time by 85%, averaging to 3 minutes.
Led data mining & cloud automation strategies, building real-time monitoring dashboards in Power BI.

AbbVie

Graduate Student Researcher - Data Mine Purdue

West Lafayette, IN, USA Jan 2024 – Apr 2024

Engineered multi-touch sequence-based attribution models using XGBoost with SHAP values for interpretability, optimizing marketing spend and maximizing ROI for pharmaceutical campaigns.
Deployed a Streamlit app for real-time analysis.

Purdue University

Graduate Research Assistant

West Lafayette, IN, USA Sep 2023 – Jan 2024

Applied data analytics and image processing tools (SOAX, OpenCV, NumPy, Pandas) to analyze actin microfilaments.
Engineered algorithms, developed performance metrics, and created a phenotype-specific registry for diverse plant types.

Labellerr

Software Developer (ML)

Chandigarh, India Jan 2023 – May 2023

Developed the Autolabel Jobs feature for automated dataset labeling using domain-specific and zero-shot models, boosting annotation speed by 82%, and integrated it into the product demo for customer engagement.
Implemented and deployed models including CLIP, DINOv2, YOLOv8, and semi-supervised learning techniques to reduce manual labeling across large-scale vision datasets.
Contributed to data engineering, model evaluation, competitive analysis, and technical knowledge sharing through presentations, blogs, and cross-functional collaboration with marketing.

Defence Institute of Advanced Technology

AI developer - Contract Project

Remote, India Oct 2022 – May 2023

Developed automated AI pose estimation model for learning and correcting multiple army drills exercise, using OpenCV, Mediapipe, and YOLOv8 for underprivileged area with no resources/internet connection.
Deployed a real-time offline pose-tracking and feedback system with automated pose correction guidance report and video snippets.
Designed and Created UI/UX using Figma and user-friendly web application using Flask.

Technological University of the Shannon

Automation Developer

Remote, Ireland Aug 2022 – Jan 2023

Built a data extraction pipeline for Enterprise and Hertz invoices using OCR, NLP, and machine learning, achieving 98% accuracy and reduced manual entry time
Outperformed RPA tools such as BluePrism and UIpath with improved anomaly detection and 75% faster processing

Digital Darwin

Data Scientist

Remote, India Jun 2022 – Aug 2022

Contributed to the development of the company's core AI-based product, executing end-to-end pipelines from raw data collection to on-device model deployment with optimization.
Worked on Pose Estimation and Image-based models, ensuring efficient training and inference tailored for real-world applications.
Participated in an NVIDIA-powered startup hackathon and established a long-term strategy for data collection and in-situ dataset curation.

Apna

Data Analyst

Remote, India Nov 2021 – Feb 2022

Built interactive internal dashboards using Retool, integrating tools like Metabase and Twilio to streamline operations and communication.
Analyzed user response data via Google Sheets and SQL to inform product and design decisions.
Created clear, insightful visualizations using Plotly.js to support data-driven decision-making across teams.

Portfolio

Multi-touch attribution modelling

Affiliated with AbbVie

Machine Learning Attribution Modelling Streamlit XGBoost Shapely values Marketing Analytics
- Developed a multi-touch attribution model for pharmaceutical marketing using XGBoost and SHAP, enabling transparent, data-driven insights.
- Built an interactive Streamlit dashboard for real-time analytics and actionable marketing recommendations.
- Optimized marketing spend allocation across multiple channels, improving campaign ROI by 18%.
- Empowered marketing teams to make informed decisions with interpretable ML outputs and clear visualizations.
Live Demo
Few-shot Defect Classification in Manufacturing Setup (PoC)

Affiliated with L&T Technology Services

Machine Learning Anomaly Detection Computer Vision Edge AI Data Engineering
- Implemented few-shot learning for defect detection with limited training data
- Designed edge AI solution for real-time manufacturing quality control
- Built robust anomaly detection system for various defect types
- Created scalable data engineering pipeline for manufacturing data
Live Demo
Pairup: Find a perfect project partner

Copyrighted application for project peer recommendation

KNN K-nearest neighbours PCA React Native
- Implemented KNN algorithm for skill-based partner matching
- Built React Native mobile application for cross-platform compatibility
- Utilized PCA for dimensionality reduction and improved matching accuracy
- Developed user-friendly interface for seamless partner discovery
View Code
ArmyPose : Validating drills using pose estimation

Ministry of Defence, Government of India

Computer Vision Deep Learning Pose Estimation
- Built real-time pose estimation system for military drills
- Implemented automated feedback and correction system
- Created offline-capable solution for remote training areas
- Designed user-friendly interface for military personnel
View Code
AutoFis: Automatic fish detection and recognition

Smart India Hackathon 2022 Winner

Deep Learning Computer Vision Online Learning ML Pipelines ONNX deployment
- Won Smart India Hackathon 2022 with innovative fish detection solution
- Implemented online learning for continuous model improvement
- Built efficient ML pipeline for real-time processing
- Deployed optimized ONNX model for production use
View Code
Healthcare appointment scheduling app

Patient-doctor appointment booking system

Node.js React.js Express.js MongoDB Heroku
- Developed full-stack web application using MERN stack
- Implemented secure user authentication and authorization
- Created responsive design for mobile and desktop users
- Deployed application on Heroku for public access
View Code
Student Mental Health Analysis

Purdue Stats Course Project

Machine Learning Linear Regression Data Analysis Statistics
- Performed extensive data analysis on student mental health dataset
- Built linear regression models for mental health prediction
- Identified key factors influencing student mental well-being
- Created comprehensive statistical analysis report
View Code
Comparison of Machine Learning Algorithms for Twitter Sentiment Analysis (IEEE)

Atharva Parikh, Riddhi Pawar, Priya Shelke, Rohit Gadhave, Jayshree Bagade

Machine Learning Sentiment Analysis Twitter Neural Networks Deep Learning
- Compared traditional ML algorithms with neural network approaches
- Analyzed performance metrics across different sentiment analysis techniques
- Provided insights into optimal algorithm selection for social media analysis
- Published findings in IEEE conference proceedings
View Paper
Offer and Deal-Quality Prediction using Machine learning and Fuzzy approach: A Shark Tank India Case Study (ACM)

Atharva Parikh, Shreya Jain, Parikshit N Mahalle, Gitanjali Rahul Shinde

Artificial Neural Networks Computational Model Fuzzy Logic Machine Learning Shark Tank India Startups
- Combined fuzzy logic with neural networks for investment prediction
- Analyzed Shark Tank India dataset for startup success factors
- Developed computational model for deal quality assessment
- Published in ACM conference proceedings
View Paper
Revolutionizing Physics: A Comprehensive survey of Machine Learning Applications (Frontiers Physics)

Rahul Suresh, Hardik Bishnoi, Artem V. Kuklin, Atharva Parikh, Maxim Molokeev, R. Harinarayanan, Sarvesh Gharat, P. Hiba

Physics Machine Learning Neural Network Deep Learning Artificial Intelligence
- Surveyed machine learning applications across multiple physics fields
- Analyzed neural network implementations in physics research
- Identified emerging trends in AI-physics integration
- Published in Frontiers in Physics journal
View Paper
Empowering India's Climate Action: Harnessing Blockchain for Carbon Trading

Shreya Jain, Atharva Parikh, Shruti Jawale, Riddhi Pawar

Blockchain Carbon Trading Decentralization Environment Sustainability
- Designed blockchain-based carbon trading platform architecture
- Analyzed decentralization benefits for climate action
- Proposed framework for India's carbon trading ecosystem
- Published in IEEE conference proceedings
View Paper
Enhanced Cognitive Quality Inspection in Manufacturing through Siamese Few-Shot Learning

Yog Dharaskar, Atharva Parikh, Ketaki Aloni, Vipulkumar Rajput

Defect Classification Anomaly Detection Manufacturing SWIN Transformer Siamese Networks
- Implemented Siamese networks for few-shot defect classification
- Integrated SWIN transformer architecture for enhanced feature extraction
- Developed cognitive inspection system for manufacturing quality control
- Published in IEEE conference proceedings
View Paper