Who I Am
Software Engineer, open-source enthusiast, and perpetual learner
What Drives Me
Musicophile, poetry & art seeker, day dreamer
Peripatetic • Introverted • Empathetic • Melancholic • Emotive
About
Education
Bachelor of Engineering
CGPA: 7.50 / 10
- 👋 Hey there everyone! I'm Mohammad Shahid, a versatile full-stack software developer with a deep passion for crafting intuitive, high-performing web applications. I thrive on building innovative solutions using modern tech like React, Next.js, Node.js and I'm deeply invested in the exciting realm of GenAI.
- I'm actively working with cutting-edge frameworks like LangChain, LlamaIndex, and agentic AI technologies including LangGraph with React and Next.js, to deliver full-stack solutions powered by GenAI.
- If you are looking for a passionate Software developer who thrives in challenging environments and is dedicated to continuous learning and innovation, I would love to connect with you. Let's create something amazing together! ✉️😊
Skills
Programming Languages
JavaScriptTypeScriptGolangPython
Backend
Node.jsExpress.jsREST APIsGraphQLOAuth 2.0Web SocketsWebRTC
Databases
MySQLPostgreSQLMongoDBRedisSupabaseFirebaseORMs
Frontend
React.jsRedux ToolkitReact QueryNext.jsTailwind CSSShadcn
DevOps
GitLinuxDockerK8sCI/CDJenkinsTerraformAWSGCPVercel
GenAI
PythonLangChainLlamaIndexVector DatabasesRAGLLM APIs (OpenAI, Anthropic)AI AgentsLangGraphCrewAIPrompt EngineeringKnowledge Graphs
Experience
Fullstack Engineer · Remote
Weya.ai | AI Workflow Automation for BFSIApr 2025 - Aug 2025
- Engineered a multi-prompt calling agent builder using a node-based visual editor, enabling conditional conversational flows with branching logic (call pickup, hang-up paths), multilingual support, and real-time state synchronization.
- Designed and implemented a visual workflow orchestration system using React Flow, enabling users to design, edit, and manage multi-agent workflows (Phone, WhatsApp, Email) via drag-and-drop UI, third-party tool integration and cross-agent coordination.
- Built real-time analytics dashboards with D3.js visualizing call volume, success rates, and latency through interactive filtering and time-series aggregation for live agent performance monitoring.
- Optimized Next.js performance using React Query caching, route-based code splitting, React.lazy() with Suspense, and ISR, reducing Time to Interactive (TTI) by 40%.
- Architected multi-factor authentication with OAuth 2.0 (@react-oauth/google), TOTP verification, JWT access/refresh token rotation with HttpOnly cookies, and role-based access control (RBAC) middleware.
- Resolved Next.js crashes caused by client-side pagination when processing 100k+ agent call records by implementing server-side pagination using MongoDB aggregation pipelines, reducing client-side memory footprint by 75%.
- Automated voice pipeline using Python and ElevenLabs API for TTS generation across multilingual agents, integrated AWS S3 with SSE-S3 encryption, and developed Node.js API with AWS SDK for pre-signed URL generation.
- Played a key role in developing distributed microservices in Go, using Azure Queue Service for message brokering and Redis for state management, implementing a weighted round-robin scheduling algorithm for real-time multi-agent calls.
Next.js Developer · Upwork
HeyMilo AIDec 24 - Jan 25
- Integrated Daily.co and Vapi SDK for video conferencing features, enhancing the platform’s video interview capabilities.
- Leveraged Vapi’s iframe object to create a seamless connection to Daily.co rooms, ensuring synchronization between the AI assistant and video conferencing.
- Utilized TypeScript for type safety and enhanced code maintainability, ensuring fewer runtime errors.
- Designed and implemented responsive and accessible user interfaces using Next.js, ensuring a consistent experience across devices.
- Optimized existing components for better performance and maintainability, reducing loading times by 25%.
Machine Learning Trainee
ThinkNEXT Technologies Private LimitedJuly 22 - Aug 22
- Employed advanced Machine Learning algorithms to predict automobile prices based on crucial factors.
- Trained a Linear Regression model and leveraged feature engineering techniques and popular libraries like NumPy, Pandas, and Matplotlib, achieving a nuanced understanding of how different vehicle attributes impact their market price.
- Planned future enhancements to refine the model and explore additional features for real-world deployment.
OpenSource
TalkRoom - Real-Time Chat Application
Live DemoFull-Stack Project
- Built a modern and secure real-time chat application with end-to-end encryption using React, TypeScript, Node.js, and WebSocket technology.
- Developed a responsive UI with Material UI, featuring protected routes and secure state management with Redux Toolkit.
- Integrated WebRTC for encrypted video/audio calls and secure file sharing with validation.
- Utilized Redis for secure session management and PostgreSQL with Prisma ORM for encrypted data storage.
- Implemented comprehensive security features including brute force protection, XSS prevention, CSRF validation, and SQL injection protection.
Campus360 - AI-Powered Virtual Assistant
Live Demo🤖 AI-RAG Project
- An AI-powered chatbot designed to assist students, faculty, and visitors at the Model Institute of Engineering Technology (MIET).
- Implemented a cutting-edge tech stack: React.js, TypeScript, Tailwind CSS for frontend; OpenAI API, Langchain.js, Firebase, Supabase Vector Database as a BaaS.
- Initiated the project by aggregating approximately 16,000 lines of institutional data from the official MIET website. This data was then extracted, cleaned, and prepared for integration with a Vector database, facilitating vector embeddings operations.
- Optimized response generation and language modeling using RAG (Retrieval-Augmented Generation), Prompt Engineering, enhancing response accuracy by 90% for common queries
- Achieved 24/7 availability, scalability, and multilingual support, significantly improving campus information accessibility and user experience.
📄Research Paper: AI-BASED CHATBOT FOR EDUCATIONAL INSTITUTES
🚀 Over 1000 reads!⭐ Like it?
MindVault
Live DemoPersonal Knowledge Management System
- Developed MindVault, a personal knowledge management system designed for researchers and academics to organize, manage, and collaborate on research papers, articles, books, and notes.
- Implemented user authentication and authorization using JWT, ensuring secure access and management of user data.
- Developed a feature for uploading, updating, and deleting research papers and notes, with support for tags and metadata for easy categorization.
- Enabled multimedia support by integrating React-Player, allowing users to watch educational videos directly on the platform.
- Developed with a responsive design, ensuring seamless usage across desktop and mobile devices.
Interviewer
Live Demo🎙️ Voice-Based AI Project
- Developed an interactive web application that simulates job interviews using AI-powered voice conversations, providing a realistic interview experience with real-time voice interaction and feedback.
- Built with Next.js 14, React, and TailwindCSS, featuring state management through React Context and API protection with in-memory rate limiting.
- Integrated OpenAI GPT-4 for interview logic and Deepgram for Speech-to-Text and Text-to-Speech capabilities, enabling natural conversations and dynamic responses.
- Implemented comprehensive accessibility features following WCAG guidelines, including keyboard navigation, screen reader support, and visual accessibility enhancements.
Owais Yaqoob
Portfolio WebsiteFreelance Project
- Led the full-stack development of a dynamic portfolio website using Next.js, Tailwind CSS, and TypeScript.
- Enhanced online presence and user engagement by 65%.
- Collaborated closely with the client to define project requirements, achieving a 100% alignment with personal branding goals, and led the end-to-end deployment process, integrating Firebase for seamless user authentication and Razorpay for secure payment processing.
- Optimized website performance through efficient code and best practices, achieving fast load times and smooth navigation.
- Integrated email.js for a robust contact section, facilitating easy and reliable communication. - Collaborated closely with the client to define website requirements.
- Ensured the final product aligned perfectly with personal branding and professional goals.
- Provided ongoing support and maintenance, including regular updates and feature enhancements.
mohammadshahid.me
Live DemoPersonal Portfolio
- Utilized cutting-edge technologies such as Next.js, Tailwind CSS, and TypeScript to develop a sleek and modern portfolio website.
- Employed TypeScript for type safety and code integrity, reducing bugs by 30% and enhancing maintainability.
- Leveraged Tailwind CSS utility classes to create a highly responsive and mobile-friendly layout.
- Optimized the portfolio for various screen sizes and devices.
- Incorporated interactive features such as smooth scrolling, parallax effects, and lazy loading to create an immersive and memorable user experience.
Agentic AI
Open-Source Repository🤖 AI Agents
- Curated an open-source repository with hands-on agentic AI projects, garnering 37 stars and 18 forks from the developer community.
- Implemented autonomous agent examples demonstrating agent-centric workflows, multi-agent collaboration, and tool-use patterns.
- Leveraged industry-leading frameworks including OpenAI, Phidata, LlamaIndex, Claude, CrewAI, LangGraph, Google Gemini, and Groq API.
- Showcased expertise in building intelligent agents with advanced reasoning capabilities, multi-tool integration, and scalable deployment patterns.
My Latest Blog Posts
Why RAG Outshines Fine-Tuning in LLM Optimization
Sep 17, 2024 | Updated: Sep 18, 2024
Understanding Why RAG Outshines Fine-Tuning in LLM Optimization.
Read More
Mastering Fine-Tuning: A Journey Through Model Optimization
Sep 5, 2024 | Updated: Sep 8, 2024
Understanding how Fine-tuning is key for optimizing small LLMs, and why mastering prompt engineering is equally vital for driving AI model behavior.
Read More
Generative AI and Agentic AI for Software Development
Jan 5, 2025 | Updated: Jan 8, 2025
This blog explores why now is the golden time to dive into these transformative AI technologies, how software engineers can leverage them, and the cutting-edge frameworks and tools shaping the future of AI-powered software development.
Read More
Introduction to Prompt Injection Vulnerabilities: What They Are and How to Protect Your LLM Applications Against Them
Feb 11, 2025 | Updated: Feb 12, 2025
A detailed introduction to prompt injection vulnerabilities, discussing what they are and strategies for protection.
Read MoreWhat People Say on Upwork
Built 4 step workflow for realtime conversation (Audio/Video).
Shahid made my portfolio the best I could expect. The color, font, and background are professional and exactly what I wanted. I appreciate his consistency and patience throughout the project, as he worked on it for a long time to ensure it was perfect. Thank you!