Ahmed is available for hire
Hire AhmedAhmed Fayyaz Butt
Verified Expert in Engineering
Full-stack Developer
Lahore, Pakistan
Toptal member since October 26, 2026
Bio
Ahmed Fayyaz has spent 8+ years building production systems and agentic AI products end to end across SaaS, CRM, and ERP platforms. He works mainly in Laravel and Python on the back end, and in Vue and React on the front end. His toolkit centers on AWS, Docker, and Linux. Ahmed designs the data layer, queues and jobs, APIs, and cloud infrastructure behind multi-agent systems, RAG pipelines, and LLM features.
Portfolio
WIZniche
Nuxt.js, Laravel, LangChain, Model Context Protocol (MCP), REST, GraphQL...
EZ-AD
LangChain, Retrieval-augmented Generation (RAG), LangGraph, ElevenLabs TTS...
Eccountant
Laravel, PHP, MVC architecture, Redis, Full-stack, Architecture...
Experience
- PHP - 8 years
- Laravel - 8 years
- Python - 5 years
- Node.js - 5 years
- LangGraph - 4 years
- LangChain - 4 years
- Model Context Protocol (MCP) - 4 years
- Nuxt.js - 3 years
Work Experience
Senior Full-stack Developer
2023 - 2026
WIZniche
- Drove end-to-end architecture and delivery of wizniche.com, a multi-module SaaS platform using Nuxt.js, Laravel, and Vue, encompassing CRM pipelines for lead management and ERP workflows for inventory and financial reporting.
- Reduced manual sales follow-up effort by 60% by designing LangChain AI agents for automated data enrichment and intelligent next-action suggestions embedded directly into CRM workflows.
- Implemented Model Context Protocol (MCP) server integrations so AI agents could call internal CRM and ERP tools in a structured, context-aware way, increasing reliability across all agentic workflows.
- Architected RESTful and GraphQL APIs with OAuth 2.0 and JWT; developed real-time dashboards via WebSocket with Pinia state management, enabling distributed teams to collaborate on live data.
- Designed serverless AWS infrastructure using Lambda, S3, RDS, DynamoDB, CloudWatch, and CodePipeline/CodeBuild, delivering zero-downtime CI/CD automation across all environments.
- Improved query performance by 40% by introducing a Redis in-memory persistence layer across high-traffic ERP and CRM operations, reducing average response time at peak load.
Technologies: Nuxt.js, Laravel, LangChain, Model Context Protocol (MCP), REST, GraphQL, JSON Web Tokens (JWT), WebSockets, Pinia, AWS IoT, AWS Lambda, Amazon S3 (AWS S3), Amazon RDS, Amazon DynamoDB, Amazon CloudWatch, AWS CodePipeline, CodeBuild, Redis, Agentic AI, Full-stack, APIs, Agentic Workflow Design, Architecture, Large Language Models (LLMs), Amazon Web Services (AWS)
Full-stack Developer
2021 - 2023
EZ-AD
- Served as the sole architect and developer of EZAI, a production AI conversational platform with multi-turn dialogue, LangChain RAG over business documents, LangGraph multi-agent workflows, and ElevenLabs TTS voice interaction.
- Authored MCP-compatible tool interfaces enabling EZAI agents to interact with internal APIs and databases in a standardized way, making the agent system extensible and production-ready.
- Engineered ezadtv.com, a digital advertising and media management platform using Node.js, Vue, and a custom PHP MVC framework, handling high-volume ad delivery.
- Implemented a real-time chat system with Node.js event-driven architecture and WebSocket, with Vuex state management for message sync and typing indicators, handling over 1,000 concurrent connections reliably under load.
- Designed GraphQL schemas and resolvers for a POS system, standardizing data queries and reducing over-fetching, resulting in measurably faster client load times.
Technologies: LangChain, Retrieval-augmented Generation (RAG), LangGraph, ElevenLabs TTS, Node.js, WebSockets, Vuex, GraphQL, Agentic AI, Full-stack, APIs, Agentic Workflow Design, Architecture, Large Language Models (LLMs), PostgreSQL, FastAPI, React
Software Developer
2017 - 2021
Eccountant
- Delivered a full enterprise ERP (eccountant.co) and POS platform (epos.pk) in Laravel/PHP with Vue using MVC architecture, Eloquent ORM, and RESTful API design, serving hundreds of active business users.
- Developed end-to-end accounting modules spanning sales management, purchase orders, chart of accounts, voucher processing, inventory management, and financial reporting with unit and integration tests validating critical financial calculation paths.
- Reduced database overhead by 40% by introducing a Redis in-memory layer for session management, transaction data, and frequently queried records in high-volume financial workflows.
- Scoped, estimated, and delivered multiple feature modules on tight timelines, building strong ownership habits and direct communication with business stakeholders.
Technologies: Laravel, PHP, MVC architecture, Redis, Full-stack, Architecture, Amazon Web Services (AWS)
Project History
EZ-AI
https://ezai.ioI architected a production conversational AI platform that runs RAG pipelines over business documents, with multi-agent LangGraph orchestration and ElevenLabs voice TTS. I built it end to end, and it's live and publicly accessible.
WIZniche
https://www.wizniche.comI built a multi-module SaaS CRM/ERP platform with embedded agentic AI workflows. I was the sole technical lead and handled architecture, delivery, and maintenance. A Redis and async queue layer cut database overhead by 40%, and the LLM agents cut manual sales follow-up efforts by 60%.
GraphOS
https://github.com/ahmedbutt2015/graphosI engineered an open-source governance and observability layer for LangGraph agents. It has policy enforcement (LoopGuard, BudgetGuard, MCPGuard), a live graph dashboard with time-travel replay, and tracing with SQLite persistence. It's published on npm and PyPI and used by the community.
Spine
https://github.com/ahmedbutt2015/spineI contributed to a CLI that turns an unfamiliar codebase into a verified onboarding guide. It detects entry points and extracts architecture from static-analysis edges. It generates a Mermaid diagram, a recommended reading order, and subsystem summaries, and serves as a Claude Code skill for /map and /onboard.
Eccountant ERP
https://eccountant.coI built and maintained an enterprise ERP and POS platform for four years, serving hundreds of active business users and integrating with key systems and APIs. I developed the solution in PHP and MySQL, owning the full lifecycle from database design and back-end logic through to deployment, support, and ongoing feature delivery.
Education
2013 - 2017
Bachelor's Degree in Computer Science
University of Lahore (UOL) - Lahore, Pakistan
Skills
Libraries/APIs
Node.js, Vue 3, React, Vuex, OpenAI API, Vue
Tools
Amazon CloudWatch, Claude Code
Languages
PHP, TypeScript, Python, GraphQL
Frameworks
Laravel, Nuxt.js, LangGraph, JSON Web Tokens (JWT), Laravel Livewire, Nuxt 3
Paradigms
REST, Model Context Protocol (MCP), Microservices
Storage
MySQL, Redis, PostgreSQL, Amazon S3 (AWS S3), Amazon DynamoDB, Database Management, MongoDB, Neo4j
Platforms
Docker, Amazon Web Services (AWS), AWS IoT, AWS Lambda, Azure, Linux
Other
Full-stack, Software Architecture, FastAPI, LangChain, Multi-agent Systems, Agentic AI, APIs, Agentic Workflow Design, Architecture, Large Language Models (LLMs), WebSockets, Pinia, Amazon RDS, AWS CodePipeline, CodeBuild, Retrieval-augmented Generation (RAG), ElevenLabs TTS, MVC architecture, Software Engineering, GCP, Cloudflare, CI/CD Pipelines, Tool Calling, Prompt Engineering, Pinecone, FAISS, ElevenLabs Solutions, Static Analysis, Streaming
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring