Hitesh Taneja
AKA. Happy :)
I’m a postgraduate Computer Science student based in London, specializing in Machine Learning and Artificial Intelligence. Currently, I’m delving deep into my MSc project, “Cluster Interpretation via Dimensionality Reduction,” exploring innovative ways to make sense of complex data.
Originally from India, I’ve always been passionate about using technology to create useful, impactful solutions for the internet. Whether it’s data analysis, web development, or software engineering, I love turning raw data into insights that can change the world.
When I’m not immersed in research or coding, you can find me sharing my experiences and learnings on Medium, beatboxing, or exploring the outdoors through traveling and hiking. I also enjoy spending quality time with family and friends, listening to music, and diving into self-help books.
Welcome to my digital space—a place where data meets creativity, and every project is a step toward making a positive impact.
- LinkedIn @hiteshtanejaa (opens in a new tab)
- Twitter @hiteshtanejaa (opens in a new tab)
- GitHub @hiteshtanejaa (opens in a new tab)
- Scaler @Hitesh Taneja (opens in a new tab)
- Email hiteshtanejaa@gmail.com
Work
So if are till here I think you are interested to know what I WORK on:
What I've been working on
I am diving into the intricacies of unsupervised learning by focusing on cluster interpretation via dimensionality reduction. The project explores the challenge of making sense of high-dimensional data by projecting it into a lower-dimensional space without losing the core characteristics that define distinct clusters. I am developing novel methodologies that not only preserve the structure of the data but also provide intuitive visual representations for interpretation. By leveraging advanced techniques such as t-SNE and UMAP alongside traditional approaches like PCA, I aim to uncover hidden patterns that conventional clustering methods might miss.
This work is especially crucial in applications where understanding the underlying data distribution can lead to better decision-making, such as in customer segmentation, anomaly detection, and bioinformatics. The research involves rigorous experiments on diverse datasets, rigorous evaluation metrics, and iterative refinement to ensure that the reduced dimensions accurately reflect the intrinsic relationships within the data.
Ultimately, this project seeks to bridge the gap between complex data analytics and human interpretability, enabling data scientists and stakeholders to derive actionable insights with greater clarity. It’s an exciting journey of blending theory with practical solutions that demonstrate the power of artificial intelligence in transforming raw data into meaningful, visual narratives.
In 2022, I joined Accenture as a Data Analyst, and then progressed to become an Associate AI Engineer, where I had the opportunity to collaborate on a critical project for an Irish financial institution. Our mission revolved around supporting the bank’s strategic closure of 63 branches — a massive initiative that required intelligent automation, secure data handling, and smart analytics.
My role was focused on designing and deploying AI and Machine Learning solutions that turned complex financial data into actionable insights. I developed RAG-based AI applications using LangChain, LangGraph, and FastAPI, allowing executives to query data conversationally and access instant insights — cutting analysis time by nearly 80%.
I also worked on forecasting and anomaly detection models built with Scikit-learn and PyTorch, and deployed them using TensorFlow Serving on AWS, improving credit-risk prediction accuracy by 13%.
Beyond model development, I engineered end-to-end ML pipelines on AWS (S3, EC2, Glue, Lambda), containerized with Docker and orchestrated via Kubernetes. I also created REST APIs for real-time model serving, integrated monitoring and logging systems, and applied CI/CD and Git-based version control to ensure reliability and scalability.
Working in a highly regulated environment taught me a lot about explainable AI, compliance, and stakeholder collaboration — combining technical precision with business understanding to deliver trustworthy AI systems.
Before Accenture, I worked at Landis+Gyr as an Engineer on their Meter Data Management System (MDMS) project — the digital backbone for smart energy analytics.
I built and optimized SQL queries (joins, subqueries, window functions) to process large-scale meter and consumption data, boosting query performance by 40%. I automated data cleaning and transformation workflows using SQL and Excel, which improved reporting efficiency by 30%, and implemented data validation pipelines that reduced operational data errors by 25%.
To bring data to life, I created interactive dashboards in Tableau* that visualized KPIs like meter reliability, consumption trends, and outage analytics — empowering business teams with data-driven decisions.
This experience helped me strengthen my foundation in data engineering, performance optimization, and visualization, forming a strong base for my journey into machine learning and AI-driven analytics.
Prior, I was working as a Tutor at StudyPool and also as Freelance Web Developer. I worked with 3 small business to help them bring their business online.
if you ever liked working with me or my work or just me being me
© Made with ❤️ Hitesh Taneja.you can consider buying me a coffee