Available for select opportunities
Fayez
Full Stack Developer with 4+ years building scalable web, mobile, and AI-driven applications — from production RAG pipelines to LangGraph-orchestrated agents to SaaS platforms shipped end to end.
About
Building the systems behind reliable AI products
A snapshot of the experience, technologies, and delivery habits behind every project below.
Full Stack Developer with 4+ years of experience building scalable web, mobile, and AI-driven applications. Proficient in React.js, Next.js, Node.js, Python, FastAPI, and TypeScript, with hands-on expertise in LangChain, LangGraph, Retrieval-Augmented Generation (RAG), and multi-LLM integration for production-grade agentic AI systems. Skilled in building RESTful APIs, SaaS products, cross-platform mobile apps, and secure authentication systems. Experienced with Docker, GitHub Actions CI/CD, Azure Container Apps, and AWS EC2.
Experience
4+ yrs
Shipping production web, mobile, and AI systems
AI Systems
Agentic
LangChain & LangGraph orchestration, RAG, multi-LLM routing
Full Stack
End-to-end
React.js, Next.js, Node.js, FastAPI, PostgreSQL, MongoDB
Delivery
Cloud-native
Docker, GitHub Actions CI/CD, Azure Container Apps, AWS EC2
Experience
Where the pipeline ran in production
Each role below fed into the next — from RESTful platforms to agentic AI systems running in enterprise environments.
Senior Software Engineer
Recro · Client: C5i
- Delivered full stack features across multiple enterprise products, building responsive frontends with React.js, Next.js, ShadCN, Tailwind CSS, and Zustand, and scalable backend services with Node.js, Express.js, and Python FastAPI.
- Designed and built agentic AI workflows using LangChain and LangGraph to orchestrate multi-step reasoning, tool calling, and stateful conversation flows across multiple AI-driven applications, improving query resolution accuracy and response reliability.
- Built and optimized RAG (Retrieval-Augmented Generation) pipelines in Python using FastAPI, integrating vector search and document chunking strategies to ground LLM responses in enterprise knowledge bases across different product lines.
- Established end-to-end CI/CD pipelines using GitHub Actions for automated build and deployment to Azure, reducing deployment time by 60% and eliminating manual release overhead.
- Containerized frontend and backend services using Docker and deployed on Azure Container Apps and Azure Container Registry, ensuring high availability and environment consistency across development, staging, and production.
- Collaborated with cross-functional teams to define technical requirements, design system architecture, and implement best practices for code quality and performance optimization.
- Added application monitoring and logging using Azure tools and centralized dashboards, enabling faster incident resolution and proactive performance optimization in production environments.
Software Engineer
DataMantis.ai
- Built a production-ready SaaS platform from scratch using Next.js, React.js, Tailwind CSS, and ShadCN, delivering a responsive, high-performance user experience across all devices.
- Engineered a multi-LLM AI chatbot using the Vercel AI SDK with dynamic switching between GPT-4, Claude, and Gemini, integrated with RAG pipelines and Pinecone for intelligent, context-aware document search and retrieval.
- Architected and optimized LLM prompt engineering strategies, improving AI response accuracy, reducing irrelevant outputs, and ensuring production-grade reliability across multiple language models.
- Integrated Stripe payment gateway for secure subscription and transaction processing, and implemented Supabase for authentication, real-time data synchronization, and cloud storage at scale.
- Implemented secure authentication and authorization workflows using Supabase Auth and JWT, following security best practices for user data protection and subscription-based access control.
- Maintained 80% code coverage by writing comprehensive unit and integration tests using Jest and React Testing Library, ensuring application stability across all critical modules.
Full Stack Developer
Domaincer / Hyring
- Built responsive e-commerce web applications, admin panels, and cross-platform mobile apps using React.js, Next.js, React Native, Expo, Tailwind CSS, and Material UI, delivering production-ready solutions on both Google Play Store and Apple App Store.
- Designed and built RESTful APIs using Node.js, Express.js, and NestJS, implementing JWT-based authentication, OAuth integrations, role-based access control, payment gateway solutions, and third-party API integrations across multiple client projects.
- Designed and optimized PostgreSQL and MongoDB database architectures with advanced indexing strategies, improving query performance by 50% and reducing data retrieval time significantly.
- Architected and built Hyring from scratch — an AI-powered SaaS job portal using Next.js, React.js, NestJS, and PostgreSQL, enabling employers to post jobs with fully automated AI interview workflows serving users across India and international markets.
- Developed a real-time AI proctoring system using TensorFlow.js for face and head movement detection, combined with tab-change monitoring, copy-paste prevention, and mandatory screen share enforcement to ensure interview integrity.
- Integrated OpenAI GPT for AI-driven interview question generation and AssemblyAI for automated video transcription, sentiment analysis, and speech evaluation, delivering comprehensive post-assessment reports including malpractice detection and candidate speaking level insights.
- Mentored and collaborated with a team of 4-5 developers, conducting code reviews and establishing development standards that improved overall code quality and delivery consistency.
Skills
A full toolkit, from pixels to pipelines
Categorized the way I actually use them day to day — frontend, backend, mobile, AI, and the DevOps that ships it all.
Languages
Frontend
Backend
Mobile
AI, LLM & Agentic Engineering
DevOps & Cloud
Testing
Tools & Platforms
Projects
Things I built, personal and professional
Two personal builds where I own the architecture end to end, plus two products shipped professionally.
Enterprise RAG System
A fully local, multi-format document intelligence platform
An end-to-end Retrieval-Augmented Generation system that runs entirely on local infrastructure. It ingests virtually any document type, parses it with Docling, and answers questions with a hybrid retrieval and re-ranking pipeline.
Highlights
- Ingestion pipeline built on Docling for document parsing, with image annotation
- Hybrid chunking combined with BM25 keyword search for hybrid retrieval
- Re-ranking stage to improve the relevance of retrieved context
- Supports PDF, PPT, Excel, Word docs, Markdown, audio files, and more
- Agentic orchestration with LangChain and LangGraph
- Runs fully local, with no external data leaving the machine
Claude-Style AI Workspace
A Claude-like chat product with files, web search, and multimodal model switching
A full chat application modeled on the Claude experience: attach files, ask questions over them, search the web, and switch between multimodal models, with every conversation persisted as its own session.
Highlights
- Attach files and ask questions grounded in them
- Web search integrated into responses
- Per-chat sessions with stored chat history
- Export responses as PDF
- Multimodal model switching
Finance AI Chatbot
Conversational financial insights over connected accounting data
Built from scratch: a finance-focused chatbot that connects to accounting platforms and answers financial questions with tables, charts, and suggestions, using a choice of leading LLMs.
Highlights
- Data ingestion from QuickBooks and Xero
- Authentication and authorization
- Multimodal model switching across Claude, Gemini, and GPT models
- Answers presented with tables, charts, and suggestions
- Download responses as PDF, text, or Markdown
Hyring
AI interview platform with built-in malpractice detection
Built from scratch: a platform where AI conducts interviews based on the candidate profile and job description, then delivers a scored, consolidated report to the employer.
Highlights
- AI-led interviews tailored to the user profile and job description
- Malpractice detection: tab switching, face detection, copy-paste prevention, network tab detection, and screen sharing
- Full interview recording, with both user video and screen video
- Voice transcription with AssemblyAI
- AI-assisted consolidated report with scores for employers
- Face and reaction detection with TensorFlow.js
AI & LLM Expertise
Agentic engineering, end to end
From retrieval and grounding to multi-step, tool-calling agents — the core of every AI system I ship.
Generative AI
GPT-4, Claude, and Gemini integrated in production products
LLM Applications
Multi-LLM chat products with dynamic model routing
LangChain
Chains and tool integrations for reasoning pipelines
LangGraph
Stateful, multi-step agent orchestration
RAG
Retrieval-augmented generation grounded in enterprise knowledge bases
Vector Databases
Pinecone and Qdrant for semantic search and retrieval
FastAPI
Python services powering RAG and agent endpoints
AI Agents
Multi-step, tool-calling agentic workflows
Prompt Engineering
Tuned prompts to improve accuracy and cut irrelevant output
Tool Calling
Agents that call external tools mid-conversation
Multi-Agent Systems
Coordinated agent workflows across product lines
Education
Academic foundation
Bachelor of Computer Science and Engineering
Anna University
Certifications
Formal credentials
Full Stack Developer (MERN Stack)
GUVI
- Completed comprehensive Full Stack Developer (MERN) Certification with 4+ capstone projects demonstrating real-world application development expertise.
- Built and deployed multiple production-ready projects showcasing proficiency in React.js, Node.js, MongoDB, and Express.js with industry best practices and coding standards.