Spandan Sahai - Portfolio

About Me

Dedicated 3rd-year CSE (AI/ML) student at SRM Chennai with a passion for intelligent systems, autonomous technologies, and software engineering. I enjoy building projects that combine AI, machine learning, and modern software development to solve practical problems.
My interests extend beyond traditional programming into embedded systems, computer hardware, vehicle technology, and autonomous systems. I also have experience with 3D modeling and video editing, which have helped me develop a strong blend of technical, creative, and analytical skills. I am always eager to learn new technologies and contribute to projects that bridge the gap between software, hardware, and real-world innovation.

Skills

Programming

Python Java C / C++

AI & Data Science

Machine Learning Deep Learning Natural Language Processing Computer Vision Recommendation Systems Data Analysis Feature Engineering

Frameworks & Libraries

PyTorch Keras NumPy Pandas Matplotlib Flask

Web & Backend

HTML CSS PHP MySQL API Integration Database Design

Software Engineering

Object-Oriented Programming Multithreading Concurrent Programming Real-Time Systems Git GitHub

Design & Creative

3D Modelling SolidWorks Blender Canva DaVinci Resolve Adobe Photoshop

Projects

Featured Project

Multimodal Sentiment Analyser

Year: 2026 | Tech Stack: Python, Flask, OpenCV, Transformers, Google STT API

Engineered a real-time, zero-latency emotion detection web application using Flask. Implemented decoupled background Python threads to process continuous inputs from webcams (OpenCV) and microphones (Google STT), fusing the resulting data with text sentiment via a custom probability weighting model.

Key Engineering Highlights:

Dynamic Late-Fusion Architecture: Fuses prediction probabilities from each modality at the decision level, dynamically adjusting weight vectors based on confidence scores.
Real-time Concurrency: Implemented a producer-consumer thread design in Python to circumvent frame rendering bottlenecks and API request latency, maintaining a fluid 30 FPS.
Fault Tolerance: Elegant handling of input stream dropouts, automatically rescaling modality confidence weights if audio or visual input is temporarily lost.

Other Projects

MovieLens

Year: 2026 | Tech Stack: Python, Pandas, Scikit-learn, NLTK, MovieLens Dataset

Performed text preprocessing
Implemented TF-IDF vectorization with bigram support for feature extraction
Built a content-based recommendation engine using cosine similarity

Developed an NLP-powered movie recommendation system using the MovieLens dataset. Combined and processed user-generated movie tags, transformed text into TF-IDF vectors, and generated personalized movie recommendations through cosine similarity analysis.Applied core NLP techniques such as tokenization, stopword removal, n-grams, andfeature engineering to build an effective content-based filtering model.

AirVoyager

Year: 2025 | Tech Stack: PHP, MySQL, Amadeus

Database Design
Integration of multiple technologies was challenging
Learnt API integration and database management

Engineered a full-stack flight booking web application with PHP and MySQL. Integrated real-time flight data through the Amadeus API and deployed the system on XAMPP.

Handwritten Digit Recognition

Year: 2025 | Tech Stack: Python, TensorFlow

CNN based image recognition project
Initially trained on a small dataset which resulted in low accuracy
I understood overfitting and model regularization

Developed a CNN-based digit recognition system trained on the MNIST dataset and also using TensorFlow, Keras and OpenCV. Fine-tuned with custom handwriting samples.

FlappyBirdInJava

Year: 2024 | Tech Stack: Java

Java Swing is used for realtime graphics rendering
Figuring out the method to render smoothly using Java Swing was complex
Learnt event driven programming principles and optimisation

Designed and implemented a Flappy Bird clone in Java with Java Swing as the renderer and event-driven architecture. Integrated MySQL for user authentication and high-score persistence.

Experience

Feb 2026 - Present

Founder's Office Intern - Marketing & Operations

MOVZZ (formerly MOVZZY)

Identified and evaluated transportation service providers across Chennai, building a database of potential partners to support future platform expansion.
Performed business development outreach by engaging taxi operators and mobility service providers to validate market demand and assess partnership opportunities.

May 2025 - June 2026

Media Team Member

Software Quality Assurance Community (SQAC), SRMIST

Created reels, highlight videos, posters, and social media posts for SQAC using DaVinci Resolve and Canva.
Collaborated with the club's core team to manage content flow and publish announcements.

Jun 2025 - Jul 2025

AI & Data Analytics Intern

Edunet Foundation

Analysed greenhouse gas emissions dataset to identify patterns and trends.
Built predictive models using historical data to forecast future emissions.

Aug 2023 - Mar 2024

Chassis Team Member

4ZE Racing, SRMIST

Applied computational and simulation-based approaches to optimize the Formula Student race car chassis in SolidWorks.

Hi,
my name is
Spandan Sahai