Bharath Shroff

Jack of All Trades, Master of Some

AI Engineer | Data Scientist

Full Stack Developer

About Me

My Career Experience

Results-driven professional with 5+ years of experience in AI/ML, Data Science and Engineering and 1+ years of experience in building web applications consistently leveraging cutting-edge technologies to drive innovation. Proven expertise in deploying scalable ML pipelines/solutions to Production and delivering actionable insights for global stakeholders.

With a keen eye for detail and a commitment to writing clean, maintainable and scalable code, I strive to create applications that are not only functional but also intuitive and enjoyable to use and have a data-driven approach providing direct impact on business decisions.

Education

Bachelor of Technology in Mechanical Engineering

Indian Institute of Technology, Hyderabad

Personal

When I'm not coding, you can find me travelling, playing football or Table Tennis or Badminton, doing yoga and meditation, volunteering.

Experience

Data Scientist

Mcsquared AI

Remote Contract

Sep 2024 - Present

Led the team to build a pipeline in Databricks feeding into a map view dashboard containing proximity hotspots of leads around existing business provided locations leveraging Bing Maps API and 3rd party Real world data sources.

Developing PoC for predictive solutions for data driven initiatives for clinical trial recruitments for manufacturing of new products.

AI Innovation Specialist - Finance

Trilogy, Crossover

Remote Contract

Aug 2024 - Oct 2024

Deriving Financial Insights using LLM chatbot built on React for the frontend and Express JS for the backend, which updated the RAG Vector DB upon new file uploads, reducing manual analysis time by an hour.

Reporting with automated consolidated financial reports with AI Insights including Profit and Loss (P&L) reports and Ledgers across 10 different entities saving 5-6 hours of manual effort required every month end.

Analyzed and identified over 100 AI-first tools and built PoCs to speed up accounting processes, reducing the need for building from scratch and improving efficiency.

Data Scientist

Mcsquared AI

Remote Full Time

May 2022 - Aug 2024

Deployed Machine Learning Survival model to production replacing the previous XGBoost model on Databricks using the medallion architecture capable of self re-training every month with new data and auto archive or promote to production based on the champion model using MLFlow for model versioning and evaluating the model performance based on C-score, a metric similar to Area Under the Curve (AUC score). Improved precision to 20% from 6% to narrow down sales leads by more than 80% making data-backed decisions and improve adoption of product by providing lower number but higher quality leads.

Deployed pipelines and several PoCs in Databricks to automate EDA and model complex datasets from 5 different real-world data sources and third-party API sources, ensuring the data warehouse was updated with both delta load and full load functionalities to support timely business decisions feeding into Power BI / Tableau dashboards.

Deriving insights from the data using Contours for PoC and Data Quality Management using Ontology Objects on the Palantir Foundry platform, creating dashboards crucial for global business decisions in the pharmaceutical industry.

Full Stack Developer Volunteer

Isha Foundation

Coimbatore, India

Aug 2021 - May 2022

Developed a web application using the open-source Odoo Framework built on Python, streamlining processes and digitizing multiple forms required to be filled by hand by 100s of visitors saving hours of work both for the visitors and the staff.

RESTful APIs for communication across four different ecosystems, enhancing integration and data flow.

Associate IT Consultant

ITC Infotech

Bengaluru, India

Jun 2019 - Jul 2021

Deployed end-to-end modules using Git DevOps for Continuous Integration/Continuous Deployment across the 4 stages (DEV->QA->UAT->PROD), ensuring seamless transitions and operational efficiency for MLOps.

Migrated project codebase from Scala on HDInsight Cluster to PySpark on Databricks with optimizations.

Automated daily tasks using Python (Selenium) and PowerShell scripts, including taking screenshots, verifying counts, and emailing recipients for Data Quality Monitoring, which improved accuracy and saved time.

Redesigned pipelines for performance by cleaning up the Azure Data Factory, reducing 300 redundant datasets to 4 and 30 pipelines to 3, which streamlined data processing and improved efficiency.

Created and modified Power BI dashboards to support business development by enabling data-informed decisions and providing actionable recommendations, leading to improved strategic planning.

RnD Intern

Dell EMC

Bengaluru, India

May 2018 - Jul 2018

Developed Python scripts for automated reporting, flagging approximately 100 high-priority reports daily, enhancing efficiency in report management.

Built Web Spider using scrapy for notifying outdated open-source packages which scrapes 4 different open-source websites.

Established a robust priority flow between four websites ensuring comprehensive package updates and improving data consistency.

Intern

Computer Institute of Japan

Yokohama, Japan

May 2016 - Jul 2016

Helped in improving the accuracy of multi-class Classification of emails and Achieved 70%+ accuracy.

Developed python GUI using Tkinter with Progress bar for the 3 different Training Testing, Validation stages.

Used gensim, SVM, Naive Bayes and TF-IDF pre-processor for best performance.

Skills & Technologies

Languages

JavaScriptTypeScriptPythonJavaHTMLCSSSQL

Frontend

React.jsNext.jsReduxTailwind CSSMaterial UISASSFramer Motion

Backend

Node.jsExpress.jsDjangoSpring BootRESTful APIsGraphQL

Databases

MongoDBPostgreSQLMySQLRedisFirebase

DevOps

DockerKubernetesAWSAzureCI/CDGitGitHub Actions

Tools

VSCodePostmanFigmaJiraSlack

Projects

A selection of projects that showcase my skills and experience in building web applications.

NAILS
A vibe-coded static website for NAILS Society
Next JSAceternity UICursorOracle Cloud Instance
View Project
BeingAwareMatters
A full-stack platform generating and tracking Advertisements
React NativeOpenAIFirebase
View Project
Buzzzest
1. FashionGPT Chatbot 2. FashionTech App using React Native 3. B2B Dashboard using NextJS
React NativeLangChainRAGNext.jsFirebase
View Project
Network traffic analysis
Extracting insights by transforming 6 million + Apache server logs and visualizing through plots showing traffic originates from 10 different countries.
PySparkPythonmatplotlibseaborn
View Project
Melanoma Classification
Achieved 85% AUC score in Identifying Melanoma using Convolutional Neural Network (CNN) models.
PyTorchPythonmatplotlibseaborn
View Project
Image classification of fruits
Multi Class Classification of Fruits using images, dataset used from Kaggle with 90380 annotated images. Leveraging Pretrained Machine Learning (ML) models like VGG, ResNet, AlexNet, Mobile Net for mobile deployable model.
PyTorchPythonTailwind CSSVercel
View Project
IIT Hyderabad Website
Built our college website from scratch using WordPress Templating which included integrating from over 10 departments.
WordPressPHPHTMLCSSJavaScript
View Project