Mohammad – Python, AWS, Django

A full-stack sports analytics platform created for delivering real-time tournament data, match insights, and interactive user experiences.

• Designed and deployed production-grade LLM applications using Ollama, integrating tool-calling workflows and hybrid RAG pipelines.
• Managed team composition and followed up on Jira 2-week sprints.
• Built scalable backend APIs.
• Optimized MongoDB data models.
• Implemented high-performance frontend components to support live updates and large-scale user engagement.
• Integrated AI-driven prediction and analytics capabilities, enabling simulation-based forecasting, intelligent data visualization, and advanced tournament insights.
• Deployed the solution in AWS using EC2, S3, CloudFront, Lambda, and CloudWatch.

Developed an intelligent investment research assistant based on a Retrieval-Augmented Generation (RAG) framework that integrates structured financial data with unstructured research documents to support context-aware financial reasoning. The system combined relational databases (SQLite) with the Qdrant vector database, enabling hybrid retrieval across tabular and semantic information sources. Designed end-to-end document ingestion pipelines for PDF and CSV datasets, incorporating semantic chunking, embedding generation, and vector indexing to improve retrieval quality and knowledge representation

• Implemented a Retrieval-Augmented Generation (RAG) architecture combining SQLite for tabular stock data and Qdrant vector database for semantic document retrieval.
• Built hybrid query orchestration pipelines that dynamically fused SQL-based analytics with vector similarity search to generate context-aware investment insights.
• Developed document ingestion workflows for CSV and PDF processing, including automated embedding generation, chunking strategies, and vector indexing.

A closed-loop Generative AI platform. It uses artificial intelligence, augmented reality, and robotics to improve pruning on small farms, specifically for high-value crops such as olive trees and vineyards.

• Reviewed and kept track of the work produced by peer developers.
• Built REST APIs using Django(DRF) and FastAPI for exposing data with a web client.
• Designed and deployed scalable applications on AWS using EC2, S3, RDS, and Elastic Beanstalk.
• Implemented CI/CD pipeline for AWS.

An official e-commerce platform for Bangladesh's first retail superstore chain. The site enables online ordering for a wide range of products, including fresh foods, groceries, and household essentials, with location-based delivery options.

• Built REST APIs using FastAPI for exposing data with a web client.
• Worked on the frontend with ReactJS.
• Developed the design tool with Polotno.
• Designed a way to transfer the user’s image from phone to iPad in the shop without storing it.
• Designed the event-driven architecture with one backend and multiple Raspberry Pis using Kafka.
• Implemented notification service with FastAPI and MongoDB.
• Implemented CI/CD pipeline.

Developed a scalable document management and knowledge-sharing platform supporting secure file storage, fine-grained access control, audit logging, and collaborative document workflows. Designed data management pipelines for heterogeneous document collections and implemented full-text search capabilities to enable efficient information retrieval across large enterprise repositories. Integrated AI-assisted document understanding and conversational interfaces to facilitate natural language interaction with organizational knowledge. The project provided practical experience in knowledge representation, document retrieval, and intelligent information systems, strengthening my interest in Retrieval-Augmented Generation (RAG), semantic representation learning, and multimodal foundation models

• Developed backend APIs using Python with the Django framework.
• Created the frontend using Vue.js.

Design and development of an inventory solution for warehouses, wholesale outlets, and retail outlets.

• Was responsible for the complete software development life cycle.
• Implemented a notification system for the current stock and the product’s expired date.
• Implemented a route plan for delivering products to the customers.
• Built dynamic reports.