Transforming raw data into actionable inference by designing scalable ML pipelines and deep learning models for real-world system optimization
"Building systems that don't just process information, but learn from it."
I started my journey building full-stack MERN applications. Today, I use that engineering background to ensure the machine learning models I train are actually built to be deployed, scaled, and integrated smoothly into production.
I care just as much about clean data pipelines as I do about algorithmic precision. My current work is heavily focused on predictive modeling, NLP, and designing functional RAG architectures.
Production-ready data science and engineering. A curated stack of frameworks, languages, and methodologies built for scale and high-fidelity insights.
Production-ready code.
Stats & storytelling.
Classical algorithms for tabular problems & data architecture.
Fast, reliable models.
Embeddings & NLP.
Metrics, serialization, and production patterns.
Predictive models, NLP pipelines, and scalable data architectures built for performance and real-world impact.
Hosted on Render (cold-start enabled).
First request may take
~60-80 seconds if the
service is idle. Please be patient while the infrastructure spins
up.
In-depth EDA and business insights. Uncovering the hidden narratives and structural truths within raw datasets.
Continuous learning is the foundation of data science. These are the documented milestones of my technical evolution.
Krish Naik
Learned data preprocessing, model building, evaluation, and deployment using Python, Scikit-learn, TensorFlow, and advanced NLP techniques.
HackerRank
Gained strong proficiency in writing complex queries, performing joins, subqueries, and advanced data filtering techniques.
Udemy
Built robust ML models using supervised and unsupervised learning techniques with Python and applied them in practical projects.
Most portfolios are static templates. This architecture was
custom-engineered from scratch to test the limits of browser
performance. I explicitly chose pure Vanilla JS over React to
gain absolute control over the render cycle. Managing an
11,000+ line codebase without a framework was brutal—it led to
severe layout shifts—but it forced me to isolate the game loop
and achieve 60fps stability.
It wasn't overengineering; it was craftsmanship.
My time as a MERN Stack Developer ingrained the discipline to build robust interfaces. Now, as a Data Scientist, I don't just display data—I build the logic that parses it.
Inspired by modern browser capabilities, I built my own context-aware recruiter assistant. Powered by Vector Embeddings and a custom Retrieval-Augmented Generation (RAG) pipeline, the system matches projects to queries and ensures zero-hallucination responses.
Engineered a centralized state management store to handle a "Dynamic Island" style UI. This context-aware switching reduces viewport clutter by 40%. Paired with GSAP and Lenis, motion serves structural storytelling, not just aesthetics.
"When something inspires me, I don’t wait for access. I build my own version, test the constraints, and make it better. Function implies form."
You might wonder why this portfolio feels so modern and
polished — it reflects my experience as a former full-stack
developer and my transition into Data Science.
It combines strict engineering discipline with machine
learning expertise.
Independent Development
Full-Stack Architecture
MERN Stack Developer
ML & NLP Focus
Deep Learning & Architecture
Background, availability, and my approach to applied machine learning and system engineering.
Nayan’s background in Full Stack development gave him a solid engineering foundation and an appreciation for how complete applications are built. Over time, he became more interested in the parts of systems where data informs decisions. Working with mathematics, statistics, and machine learning on real-world problems felt more engaging to him, and today he focuses on combining his engineering experience with applied machine learning to build practical, usable solutions.
Nayan is currently available to join immediately. His academic commitments are complete, and he is fully focused on building real-world ML systems and pursuing full-time roles. He is open to both remote opportunities and relocation within Pune and Nagpur, Maharashtra, India only, depending on the role and team.
Nayan’s primary focus is Applied Machine Learning. He enjoys taking models beyond notebooks and integrating them into usable products. While he keeps up with current research to stay technically sharp, his strength lies in applying those ideas to practical, high-fidelity systems.
This portfolio was designed and built entirely from scratch, not from a template. The goal was to reflect how applied ML tools are built, presented, and trusted in real-world environments — not just to look polished. This engineering-first approach is influenced by Nayan's background as a former MERN stack developer.
I am the deterministic logic behind this ecosystem. As an Applied Data Scientist, I engineer neural architectures and scalable pipelines—transforming raw chaos into high-fidelity intelligence.
> Accuracy is nothing without deployment.
Current Location
Pune Maharashtra, India
Status
Deployed
Project Completed
12 + Live Model
Portfolio
Version 2