Hi I'm Dhwani, a Frontend and AI/ML developer building intelligent and user-centric solutions.
About Me
I'm Dhwani Vyas, I'm a Full-Stack and AI/ML developer. I'ts been above 3 years since I wrote my first "Hello World". It's fun to be a programmer because it's not the same all the time. I love learning and getting adapted to new tech stack and for me problem solving is something I was born to do. Other than that in my free time u will find me reading books(my fav one being "Shoe dog" by Phil Knight, really a great book), cooking(trying and building some new flavour) and walking in nature(solus with music and trees around is just as perfect as it sounds). And yes, I am proud and love my choice of being a|
I'm open to Job opportunities where I can contribute, learn and grow. If you have a good opportunity that matches my skills and experience then don't hesitate to contact me.
My Projects
Projects I/We worked on.
AI Email Automation System
Developed a backend automation system integrating Gmail API for real-time email ingestion, with LangGraph stateful agentic workflows for multi-class intent classification and dynamic multi-step task orchestration. Implemented a Retrieval-Augmented Generation (RAG) pipeline using ChromaDB as a vector store for context-aware, semantically grounded response generation across high-volume email threads. Deployed scalable REST APIs with FastAPI and LangServe for efficient, production-ready serving of the automation pipeline.
CI/CD Pipeline to Kubernetes with Helm
Architected an end-to-end CI/CD pipeline using Jenkins multi-stage pipelines to automate build, testing, containerization, and deployment workflows across multiple environments, eliminating manual deployment overhead. Containerized microservices using Docker and orchestrated deployments on Kubernetes with Helm charts, along with shell scripts for automated image publishing, chart packaging, and versioned releases. Integrated health checks and smoke testing stages within the pipeline to ensure deployment reliability and reduce post-release defects, while managing environment-specific configurations using Helm values for staging and production.
Full-Stack Fintech Platform
Built a full-stack fintech platform featuring wallet provisioning, group ledgers, equal-split expense tracking, and real-time settlement suggestions backed by a normalized PostgreSQL relational schema spanning 4+ entity types. Integrated Stripe Checkout API for secure wallet top-ups and peer-to-peer settlement execution, alongside multi-provider authentication (email/password + Google OAuth) with Auth.js and secure session management. Designed a modular TypeScript codebase with Prisma ORM for type-safe relational data access, deployed on Vercel with optimized server-side rendering via Next.js App Router for fast, SEO-friendly page loads.
AI Resume Analyzer & Job Portal
Developed a full-stack job portal with AI-based resume analysis to match candidates with relevant job roles. Implemented NLP-based resume scoring to evaluate resumes based on skills, keywords, and job descriptions, along with ML-driven classification and job recommendations. Built secure Django-based authentication with role-based access for recruiters and applicants, and features like job posting, application tracking, and resume parsing to streamline the hiring process.
Realign AI
Built an AI-driven construction planning platform that retrieves optimal building layouts from a dataset of 571+ plans using ResNet18 embeddings. Implemented JWT-based RBAC authentication, a FastAPI backend with PostgreSQL, and an interactive Next.js frontend for visualizing and generating layout recommendations.
StockPred
Developed an end-to-end ML-powered stock prediction platform that generates directional trading signals using historical and live market data. The system leverages feature engineering and an XGBoost-based classification model, exposed via a production-ready FastAPI backend and consumed through a modern web interface.
PlantSense AI
Built an end-to-end AI-powered plant identification platform that classifies plant species from uploaded images using a deep learning model and provides structured plant care recommendations through a production-ready API.
PaperSimplify-Understanding Research Papers Made Easy
PaperSimplify is a lightweight web application that helps students and researchers understand academic papers faster by converting dense research PDFs into clear, section-wise simplified explanations using AI.
Protract: AI-Powered Fraud Prevention System
Project with which won 1st runner up at Breach25 hackathon. Developed an AI-powered fraud detection system that analyzes user behavior to identify and prevent fraudulent activities in real-time, enhancing security for online transactions. (Website is currently down due to hosting issues.)
My Certificates
Professional certificates and courses I have completed.
Get In Touch
Please contact me directly at dhwani.vyas05@gmail.com or through this form.