Problem

Executives lack real-time portfolio visibility; PMs spend too much time on fragmented monthly reporting and forecasting.

Solution

Centralizes project and org data (SharePoint, Acumatica CRM, Dayforce HR), enabling drill-down dashboards and a PM module for monthly health checks, cost forecasting, and EVM.

Tech Stack

Status

In-user testing with internal teams.

Problem

Home builder loses leads because estimates take too long and accuracy varies with experience and workload.

Solution

Ingests existing Excel templates and historical job costs, enriches with nearby comp data, and uses an LLM + RAG layer to generate a draft estimate with line items and assumptions for review.

Tech Stack

Status

Concept defined; next step is a prototype with mocked cost history and template ingestion.

Problem

Homeowners constantly ask for updates; builders waste time on repetitive status back-and-forth and sharing photos across channels.

Solution

A lightweight client-facing portal where builders post periodic updates (milestones, schedule shifts, photos/media) and costs sync in via Sage Construction Management APIs.

Tech Stack

Status

Concept defined. Next task is to create a prototype with mocked data.

Problem

Crowdfunding excludes people who can't donate cash but can contribute time and skills; volunteer coordination is messy.

Solution

Campaigns raise volunteer hours with skill matching, hour-goal tracking, approvals, search/filters, and OAuth login.

Tech Stack

Status

MVP backbone built (DB schema + auth + core flows). Frontend components in progress.

Problem

AI can meaningfully contribute to coursework, but most workflows are opaque and lack accountability and integrity controls.

Solution

Built a structured AI-collaboration framework (multiple specialized agents) that logs every interaction and ties work to evidence and git commits, backed by a CCT methodology and an ethics ledger.

Tech Stack

Status

Custom AI agents and their interaction framework are built and operational, with end-to-end logging and commit-level traceability in place. Capstone in development: an engineering knowledge preservation system that extracts key data from large documents before cloud storage deletion.

Problem

Producing consistent, high-quality AI content in Egyptian Arabic is time-heavy: trend tracking, scripting, translation, and pipeline management.

Solution

Zero-infra automation that monitors AI sources and creators, scores opportunities with Gemini, and generates English scripts then localizes to Egyptian Arabic (RTL Google Docs), orchestrated via GitHub Actions and managed through Notion workflows with notifications.

Tech Stack

Status

Fully operational system (trend scouting, creator monitoring/discovery, script writing, Arabic localization, and on-demand agent). Running on schedules and actively used for a 2–3x/week publishing cadence target within low/no-cost limits.

Problem

Creating Egyptian Arabic scripts from multiple English videos takes hours: research, synthesis, dialect translation, b-roll, and formatting.

Solution

Browser-native workflow: queue videos on YouTube, fetch transcripts, synthesize across 5–10 sources, generate a 10–12 min Egyptian Arabic script, output a properly formatted Google Doc, and log metadata to Notion.

Tech Stack

Status

End-to-end working extension (collection → transcript → generation → Google Doc → Notion) with OAuth, retries/backoff, and token expiry handling. Cuts script creation from ~3–4 hours to ~2–3 minutes.