Available for SWE & AI Opportunities

Ayush Dey

Full-Stack Software Engineer & Machine Learning Developer

I am a passionate Full-Stack Developer and AI Enthusiast with a strong foundation in modern web technologies and deep learning.

My focus is on creating highly attractive, scalable, and modular applications that bridge the gap between complex AI capabilities and intuitive user experiences.

InstitutionLovely Professional University

ProgramBachelor of Technology - Computer Science and Engineering

Academic StandingCGPA: 0.0

Email Me GitHub LinkedIn

Selected Work

Engineering Projects

FundConnectAI

Jun 2026 - Jul 2026

Full-stack crowdfunding platform

Source Live

—Architected a full-stack crowdfunding platform using React Node.js Express and MongoDB featuring a social-media-style campaign feed and analytics
—Integrated LLM APIs via OpenRouter to generate automated campaign copy and compute dynamic AI Trust Scores to evaluate campaign credibility
—Implemented secure Razorpay payment processing Cloudinary media pipelines and JWT authentication with role-based access control.

Technical Takeaways:Full-stack architecture • LLM API integration • Payment gateway integration

ReactNode.jsExpressMongoDBRazorpayCloudinaryJWTLLM APIs

Vernacular_FD_Advisor

Apr 2026 - May 2026

Multilingual financial advisory chatbot

Source Live

—Developed an end-to-end multilingual financial advisory chatbot supporting 3 Indian languages (Hindi Telugu English) and transliterated text (Hinglish) using LLMs (Groq/OpenAI) and Express.js
—Engineered a lightweight NLP pipeline combining Unicode script range analysis and keyword heuristics for real-time language detection intent classification and financial entity extraction
—Executed a deterministic calculation and simulated booking engine translating natural language financial queries into real-time Fixed Deposit return projections and structured transaction summaries.

Technical Takeaways:Multilingual NLP • Chatbot development • LLM integration • Financial computations

GroqOpenAIExpress.jsNLP Pipeline

IOT Sensor Data Compression AutoEncoder

Mar 2026 - Apr 2026

Deep autoencoder compression pipeline in PyTorch

Source

—Designed a deep autoencoder compression pipeline in PyTorch to compress 6-channel multivariate IoT telemetry (CO LPG smoke temperature humidity light) into a compact latent representation reducing transmission bandwidth and storage overhead
—Formulated an end-to-end evaluation framework utilizing MSE and MAE reconstruction loss alongside adaptive error thresholding to validate signal fidelity and preserve critical environmental variance during decompression
—Built a full-stack real-time analytics system pairing a FastAPI inference backend with a React & Recharts dashboard for batch sensor data ingestion dynamic compression benchmarking and interactive signal reconstruction visualization.

Technical Takeaways:Deep Learning autoencoders • PyTorch • Real-time data visualization • FastAPI

PyTorchFastAPIReactRecharts

Ayush Dey

Technical Expertise

Programming

AI/ML

Frontend/Backend

Databases

Tools & Tech

Soft Skills

Engineering Projects

FundConnectAI

Vernacular_FD_Advisor

IOT Sensor Data Compression AutoEncoder

Specialized Bootcamps

AI-Driven MERN Stack Bootcamp

Certifications

React.js

Basic to Beyond Python

Linux Command and Shell Scripting

Education

Bachelor of Technology

Experience

Seeking Full-Stack & Machine Learning Engineering Roles