Alok Jha
Full Stack Engineer
Full Stack Engineer
I build modern, scalable web applications end-to-end using technologies like React, Next.js, NestJS, and GraphQL. I’m experienced across the full SDLC, with a focus on writing clean, maintainable code for both frontend and backend. I also care deeply about user experience—crafting responsive, accessible UIs with Tailwind CSS, SCSS, and thoughtful design principles to ensure great usability across devices.
Once the application is built, I help clients implement robust DevOps workflows, including automated testing, CI/CD pipelines, and cloud-based deployment strategies to ensure a smooth and reliable Go-Live. I work with tools like GitHub Actions, Docker, Terraform, and NGINX to streamline development and deployment. Whether deploying to AWS, Vercel, or DigitalOcean, I focus on automation, monitoring, and repeatability to support scalable production environments.
Build applications that are scalable, resilient, and optimized for the cloud. As a certified AWS Cloud Practitioner, I bring a solid understanding of cloud architecture and best practices for designing modern, distributed systems. I have hands-on experience deploying microservices using AWS, Docker, and Kubernetes, along with infrastructure automation via Terraform and system monitoring using Prometheus and Grafana.
[will add after end of internship]
May 2025 – Aug 2025Managed Debian-based Linux systems (Red Hat and Ubuntu) on-site, conducting security compliance checks and hardware maintenance while authoring technical documentation, resulting in a 25% reduction in system downtime.
Centralized IT asset management by enrolling devices into Active Directory and Microsoft Intune, streamlining device control, patch management, and software updates across university-owned systems.
Deployed and managed endpoint security solutions such as BitLocker and device encryption, ensuring university devices remained secure against unauthorized access and data breaches.
Engineered computer vision solutions for assistive technology with Cornell Bowers CIS faculty, leveraging LLAVA, YOLO, and CogLVM frameworks to achieve 90% reduction in hallucinations and 80% increase in model accuracy.
Refined AI model deployments in Docker, boosting performance by 90% and cutting object recognition time by 40%.
Developed novel multi-modal AI system integrating visual and auditory inputs, enhancing navigation assistance for visually impaired users with 70% improvement in obstacle avoidance accuracy.
Spearheaded the development of AccessWay, a travel planning platform that uses AI and text-to-speech (TTS) technologies to create inclusive and personalized itineraries for travelers with disabilities.
Engineered intelligent systems that match user preferences and accessibility needs with real-time data on transit, accommodations, and points of interest.
Championed user-first design principles and coordinated weekly feedback cycles, ensuring the platform effectively supports solo and group travelers with diverse mobility, sensory, and cognitive needs.
A collection of my recent work and experiments
A comprehensive social media platform built with modern full-stack technologies. Features user authentication with OAuth (Google/GitHub), real-time posts with rich text editing, image uploads to AWS S3, comments and likes system, user profiles, and responsive design. Implements GraphQL and REST APIs, role-based authorization, and real-time updates. Deployed on Render (backend) and Netlify (frontend) with DynamoDB for scalable data storage.
Built a full-featured messaging app with React and Firebase (Auth + Firestore) enabling real-time chats, user auth, and clean UI. Includes robust state management and error handling.
Multiplayer word search game using Java, WebSockets, and JavaScript. Features real-time communication, score tracking, chat, and dynamic UI with game rooms and grid generation.
Real-time object detection app using TensorFlow.js, OpenCV.js, and React. Supports live camera input, image uploads, and multiple pre-trained model options for bounding box detection.
Built an intelligent document query system using Python, Chainlit, FAISS, and Hugging Face embeddings. Enables natural language search across PDFs, with 95% retrieval accuracy and fast response times.
Want to hire me or need a hand with your project? Let's connect!