A full-stack construction management platform built for an industrial contractor to replace fragmented spreadsheet-based workflows with one centralized production system.
Built a T&M estimating system with crew-based labor and cost/billable rate calculations, GPS-verified field time tracking, per-diem and cost-code allocation, purchase order and commitment tracking, cost reporting, payroll processing, client-facing billing views, and invoicing.
The platform includes 28 modules, 9 permission roles, REST APIs, and print-ready PDF/Excel exports, and was shipped iteratively while being used in daily production.
Stack: Laravel 12 · PHP 8.3 · MySQL · Alpine.js · Tailwind CSS · REST API
0
81
Thesis Evaluation Management
Designed and developed a database-backed web application for managing the thesis evaluation process at an academic institution.
The system manages student and thesis information including registration number, degree/programme, thesis title, supervisor and co-supervisor. It also manages thesis submission according to prescribed templates and verification requirements, including formatting, plagiarism and other guidelines.
Students who qualify proceed to thesis defense before a panel comprising the supervisor and internal and external members. The system maintains the status of the thesis through these stages and supports final publication after successful evaluation and prescribed formatting.
My contribution: Database design and application development.
Technology: MariaDB, C#, .NET Core, Entity Framework Core, ASP.NET (http://ASP.NET) Core Identity and RDL Reports.
0
69
QueryMind lets you query a real database using plain English.
Type a question, get an answer — no SQL knowledge needed.
Built a production-grade AI agent using LangGraph that converts
natural language to SQL, executes it on a real PostgreSQL database
with 100,000+ rows, and automatically corrects errors in a retry
loop if the query fails.
Key features:
7-node LangGraph agent with conditional routing and self-correction
pgvector semantic search for intelligent schema retrieval
Full observability dashboard with real-time metrics and query history
Trace replay — see every step the agent took to answer your question
Deployed live: Next.js frontend on Vercel, FastAPI backend on Render
Tech: LangGraph · Groq LLaMA 3.3 70B · PostgreSQL · pgvector ·
FastAPI · Next.js · Tailwind CSS
Github: https://github.com/moiz-mansoori/QueryMind-NL2SQL-Agent-System
1
292
😍
0
30
I recently engineered a real-time tracking system for a school to replace static reports with continuous feedback. Using a Google Sheets-to-Looker Studio pipeline, I automated the transition from raw teacher observations to parent-facing insights.
Key Deliverables:
Live Pipeline: Structured Sheets input for teachers to log 12+ data points per student daily.
Grit Scoring Engine: Built custom logic weighting effort over mastery using cohort benchmarks.
Dynamic URL Security: Implemented unique, filtered dashboard links for secure parent/student access.
Automated Insights: Built-in trend indicators (Improving/Stable/Declining) for instant intervention.
The Impact:
The client now uses a live command center that eliminates manual reporting and provides real-time actionable intelligence.
I developed a Python-based automation bot that monitors internet speed and logs download/upload values every 15 minutes. The data is saved in real-time to a Google Spreadsheet, along with timestamps and direct result links from Speedtest.net (http://Speedtest.net). The bot can run silently or display a GUI countdown. It helps remote teams, IT professionals, and analysts track internet stability over time.
Problem:
Financial data from stocks and crypto APIs was scattered, refreshed manually, and not ready for analytics or ML use.
Solution:
Built an Apache Airflow pipeline to collect, transform, validate, and load real-time financial data from multiple APIs into PostgreSQL, MongoDB, AWS RDS, and Qdrant.
Tools:
Apache Airflow, Python, PostgreSQL, MongoDB, AWS RDS, Qdrant, APIs
Result:
Automated sub-hourly data refresh, processed thousands of records daily, and delivered clean data for dashboards, analytics, and vector search.
0
105
Resilient Python Data Extraction Engine & Automated ETL Pipeline. Architected fault-tolerant data ingestion with session persistence, proxy rotation, exponential backoff retries, and automated schema normalization feeding directly into PostgreSQL/data warehouses.
0
65
Developed a normalized RDBMS for an automated Forex trading platform, designing interconnected tables and relational schemas to manage market data, technical indicators, ML predictions, trading signals, portfolios, execution records, risk metrics, and backtesting results.
0
33
Multi-Drive Architecture: Synchronizing & Organizing 4 Enterprise Streams
Project Overview:
The client operated with four distinct Google Drives (Personal, Giant Ventures, WMBG Rentals, and Lock & Key Property Inspection) all linked under a single account. The data was heavily fragmented, with business assets, rental documents, and personal backups mixed across all platforms. I was hired to consolidate and structure this 100GB+ ecosystem into a unified, professional workspace.
The Challenges:
Cross-Drive Clutter: Files from different businesses were intermingled, causing massive confusion and slow retrieval.
Storage Waste: High volume of redundant backups and temporary files taking up valuable cloud space.
Syncing Issues: Inconsistent folder naming prevented smooth access across mobile and desktop devices.
My Solution & Workflow:
Unified Folder Hierarchy: Designed a master structure that separated the 4 entities while maintaining a logical link between them.
Deep Data Sorting: Manually audited and moved over 100GB of files into their respective business or personal categories.
The "Zero-Duplicate" Protocol: Identified and removed redundant files to optimize storage and reduce costs.
Standardized Naming Convention: Established a consistent file-naming system for "Lock & Key" and "WMBG Rentals" to ensure future scalability.
The Results:
Turnaround Time: Completed the entire migration and organization within 3 days (ahead of schedule).
Efficiency Boost: The client can now locate any document from any of the 4 drives in seconds.
Seamless Sync: Guaranteed 100% folder accessibility across all client devices.
0
93
Built an eBay Product Scraping Web App from scratch using Python (Django) backend and Tailwind CSS with Django Templates frontend. Users can set number of processing threads, product fetch limit, and provide an eBay product list URL to scrape. The app processes listings efficiently, scraping key product details while showing live progress. Fully responsive UI, optimized for performance, and capable of handling large-scale scraping tasks with configurable concurrency.
Another successful knowledge management migration completed.
This project involved migrating approximately 200 pages across 48 Quip files into Notion before Quip's sunset. Since there isn't a reliable way to automate the migration, I manually exported and rebuilt the content while preserving structure, formatting, and hierarchy.
To make the workspace more valuable than the original, I used Notion AI to organize information, extract action items, clean up documentation, and transform scattered notes into a searchable, future-ready knowledge base.
A smooth migration with zero compromise on content quality and a much better workspace for the team going forward.