PDF chatbot with Redis-backed memory
Combined Redis and ChatGPT to store conversation history and document knowledge as embedded vectors — efficient context management with accurate retrieval-augmented answers from uploaded PDFs.

I'm a senior software engineer who designs and ships backend systems for retail commerce — platform services, partner integrations, cloud infrastructure, and the internal tooling that helps teams move faster. Backend-leaning, cloud-fluent, AI-curious.
I help you find solutions in complex software problems — from backend systems and cloud infrastructure to integrations and internal tooling that helps teams ship with confidence. Backend-leaning, cloud-fluent, AI-curious.
Most of my work sits in the integrations layer — the connections we maintain between Rewaa and partner platforms like Salla and Zid. The work spans product creation, order flow, and inventory sync; the goal is making sure each integration handles the partner's edge cases without compromising our platform's contract.
Alongside that, I've worked on improving our on-call reporting, and helped the team move toward an automation that uses Jira MCP to fetch ticket context, pull supporting data from the database, verify the underlying issue, and produce a short triage report — cutting first-pass investigation time on incoming tickets.
Currently I'm building the onboarding hub for new customers — a flow designed to get new merchants from signup to first real value as quickly as possible. The returns are multi-faceted: faster activation, fewer support touches, and a much cleaner first-30-days story for the customer.
Eonhealth
Led a comprehensive reporting module giving hospitals detailed cohort insights via Tableau, on top of AWS S3 / RDS / ECS-Fargate. Engineered a PHI scrubber for HIPAA-compliant dev access to real datasets. Migrated monolithic coreflows onto Conductor, breaking them into modular microservices for scalability and observability.
Devflovv
Built a multi-agent group chat with ChatGPT APIs, AutoGen, and FastAPI + WebSockets, using Langchain for retrieval. Shipped a web-scraping backend in Django + PostgreSQL. Built AWS cost-audit dashboards in Athena that reduced monthly cloud spend, and optimized Redshift queries and AWS Glue ETLs.
LAAM
Led MLOps using VertexAI and Elastic Search to build robust pipelines. Built a notification service in Flask + FCM + Pub/Sub. Shipped real-time product tagging on BigQuery + Dataflow. Audited GCP resources for a multi-month cost reduction of ~$2,000.
via Upwork
Built custom ChatGPT-based chatbots with memory backed by Redis as a vector store, and a CV-parsing tool using OpenAI + prompt-engineering techniques.
We Over I
Built a user-facing dashboard in React + Redux Sagas, integrating APIs from mobile services and a headless CMS.
Combined Redis and ChatGPT to store conversation history and document knowledge as embedded vectors — efficient context management with accurate retrieval-augmented answers from uploaded PDFs.
Upload a W2 PDF and automatically extract name, address, and income via OCR. Backend in Django, frontend in Next.js, structured output ready for downstream tax workflows.
FastAPI microservices powering an admin dashboard — sales/revenue insight, inventory state, and new-product registration as a microservices template.

How to run language models on your own machine inside a container, exposing them as APIs to power local GenAI development without depending on external services.
The best way to reach me is email or LinkedIn.