Ayaan Ahmad

Full‑Stack • Data Pipelines • Automation

I build lean systems that turn messy data into useful decisions.

Focused on pipelines, automation, and clean UIs. Recent work includes a manifest analyzer for ROI‑based reselling, a YouTube shorts pipeline, and community/product experiments. Fast learner, ship often.

PythonPandasFastAPI/FlaskReact/Next.jsSQLAPIsDocker

Quick Actions

TechLiquidators Manifest Analyzer

TechLiquidators Manifest Analyzer

ETL + Pricing Intelligence: ingest manifests → enrich via APIs → compute ROI + recommended max bid

  • Ingest CSV/XLSX manifests; normalize UPC/SKU/qty/condition
  • Enrich with UPCItemDB + 90‑day eBay sold comps (IQR‑trim outliers)
  • Fees/shipping/defect knobs → net resale → Recommended Max Bid
  • Sortable dashboard + CSV export; caching + retry for flaky APIs
PythonPandasStreamlitSQLiteeBay APIUPCItemDB

Impact: Turned hours of manual pricing into minutes; more consistent bids & margins

CodeDemo
YouTube Shorts Automation Pipeline

YouTube Shorts Automation Pipeline

End-to-end content ops: topic → script → clip assembly → captions → upload scheduler

  • LLM prompts → script + SEO title/hashtags
  • Clip trimming, TTS/music overlay, burned-in captions
  • Retryable job queue; idempotent uploads; progress logging
PythonMoviePyFFmpegGCP/AWS (optional)YouTube API

Impact: Batched production of shorts; consistent cadence without manual toil

CodeDemo
Masjid App (Community Events & Prayer Times)

Masjid App (Community Events & Prayer Times)

Local discovery: mosques near you, programs, sports runs, room booking, notifications

  • Geo search + activity feed; age-range event notifications
  • Basic organizer dashboard; RSVPs and interest signals
  • Future: social graph + ‘who’s going’ visibility
Flutter/React NativeSupabase/FirebaseMaps API

Impact: Activates under-served communities with a lightweight, modern UX

CodeDemo
Veyra — AI Wardrobe & Styling Assistant

Veyra — AI Wardrobe & Styling Assistant

Closet scan + try-on concepts; overlap CV + LLM for outfit suggestions & social closet sharing

  • Prototype: item detection, metadata extraction, outfit graph
  • Explores multi-shot ‘closet sweep’ vs single-item capture
  • Cost modeling for inference overlap (accuracy vs spend)
PythonOpenCVLLM APIsNext.js

Impact: Faster digitization and styling; potential marketplace hooks

CodeDemo
Paths — Skills→Role Mapping (Career Pivot Tool)

Paths — Skills→Role Mapping (Career Pivot Tool)

Enter skills/degree → get target roles, gaps, and 6‑week learning plan (YC-style MVP)

  • Role ontology + skill graph; gap analysis
  • Auto‑generate micro‑curricula and project prompts
  • Room for credential upload + verification
Next.jsPostgresLLM APIsShadcn UI

Impact: Turns vague transitions into a concrete, stepwise path

CodeDemo