A small selection of My Recent Blogs
A small selection of Recent projects
Fine-Tuning Gemma Models in Keras using LoRA
We fine tune a Google Gemma model in Keras using the LoRA technique
Multi-Agent Hedge Fund Analyst
A multi-agentic system that simulates a Hedge Fund based on 9 agents to give investment advice to the user based on stock tickers
My Portfolio Page
Latest Portfolio Page.
Fake News Detection using Graph Neural Networks
This project explores the use of Graph Neural Networks (GNNs) for fake news detection. By leveraging the graph structure of news articles and their relationships, the model captures relational information for more accurate classification.
Starbucks-Delivery Route Optimisation
Optimising Delivery Routes for Stabucks in London using Graphs and making use of Dijkstra's algorithm to solve it
Academic Research Paper Information Retrieval System
Academic Research Tool made using LangChain, PaLm-2, FAISS and front-end made using Streamlit
Uber Analytics ETL Project
A Data Engineering Project built on Google Cloud Platform and Mage Data Tool to analyze data from Uber.
Employee Churn Model Predictor
A predictive tool designed towards predicting the churn rate and enabling companies to retain their employees for longer periods using insights of the project.
Old Portfolio Page
Previous Portfolio Page.
Grocify
Grocify is a your community-focused e-commerce grocery store webapp aimed at serving the community in the best possible way
Movie Sentiment Reviews
This project aims to predict the sentiment of movie reviews using state of the art machine learning models.
Pulsar Star classification
Research project aimed at findint the optimal technique to use, Dimentionality reduction and then resampling or vice versa and then apply Machine Learning and Deep Learning models to classify pulsar stars for the HTRU dataset.
Netflix Analytics Dashboard
A Dashboard made in Tableau showcasing the various Analytical stats of British Airways ranging from most travelled places to best performing service
British Airways Dashboard
Customer Segmentation Oriented dashboard for Netflix
King County,Washington House Sales Dashboard
A Dashboard made in Tableau showcasing the various Analytical stats of King County, Washinton House Sales
Patients Analytics in PowerBI
A Dashboard made in PowerBI showcasing the various Analytical stats of a hospital with key fields being number in-patients and out-patients.
Kind words from satisfied stakeholders
During his time at DaaS he was exceptional with deliverables and new ideas. He was very active and open to new solutions. I highly recommend him for any organization. He has good Communication, technical knowledge as well as very good team coordination.I wish him very best for the future endeavours and he is always open to come work with us.
Jyotish Prasad LaheHead of Operations - DaaS
During his time with Siemens Digital Industries Software, Pragyan has consistently displayed a high level of professionalism, adaptability, and a collaborative spirit. He excels in communication, both with team members and stakeholders, and is always willing to assist others when needed. His problem-solving skills and positive attitude have made him a valuable member of our team. Pragyan is punctual, reliable, and demonstrates great responsibility in managing his tasks. He handles challenges with a calm, solution-oriented mindset, making him an asset to any environment that values teamwork and integrity.
Angus MarshallDirector of Global Sales & Customer Success - Siemens Digital Industries Software
He demonstrated great skills in his position as a Data and Business Analyst during the internship period at The Sparks Foundation. He has excellent analytical skills and his verbal communication skills are extremely organized. He can work independently, and can effectively multi-task to ensure that all projects are completed promptly. He was always willing to offer his assistance and had an excellent rapport with the many of his fellow interns. I am confident that he will devote his self to a position with your organization with a high degree of diligence. He would be an asset to any employer, and I wholeheartedly recommend him for any endeavor he chooses to pursue.
Prishita DasAssociate Data Scientist - NeoStats
As a fellow academic person, I have observed Prag's exceptional attention to detail in work reports, his open-mindedness to new ideas, and his profound curiosity to learn. These qualities make him an outstanding candidate and a valuable asset in any setting.
Patalo SelobakinConsultant - Kearney
My work experience
My approach
Planning & Strategy
Before diving into the data, we'll collaborate to define your business objectives, key performance indicators, and the role AI and data science can play in optimizing your operations. We'll map out data sources, analytical approaches, and the infrastructure needed to drive insights and automation.
Model Development & Iteration
With a well-defined strategy in place, I get to work—building data pipelines, training machine learning models, and fine-tuning algorithms. Transparency is key, so you’ll receive regular updates on performance, insights, and adjustments needed to align with your goals.
Deployment & Optimization
This is where we bring intelligence to action! The refined models and AI-driven solutions are deployed into your workflow, ensuring seamless integration and maximum impact. Post-launch, we monitor performance, fine-tune models, and scale solutions to keep your business ahead of the curve.