ScrollCollector
ScrollCollector is a privacy-first Chromium browser extension that intercepts, sanitizes, and categorizes social feeds in real time. Running entirely on-device, it leverages Chrome's built-in Gemini Nano model or local keyword filters, storing items locally without network dependencies.
The Context
Social feeds and developer blogs are high-noise, high-density environments. Curating insights manually is slow, and sending every scrolled post to a cloud LLM compromises user privacy while introducing unacceptable network latency and API costs.
ScrollCollector runs a zero-latency, privacy-preserving classification pipeline inside the browser. It runs completely locally, sanitizing PII (emails, cards, phones) before triggering on-device classification models.
What I built
- Manifest V3 Chrome Extension. Injects lightweight, scroll-optimized content scripts to capture posts using obfuscation-proof React data selectors rather than dynamic classes.
- IntersectionObserver Virtualization. Clears off-screen cards and replaces them with height-matched spacers to maintain 60 FPS scrolling and low memory usage, even with thousands of items.
- Offline HTML Diary Export. Includes a built-in exporter that compiles the collected feed stream into standalone, responsive, styled HTML diaries with interactive filters and search.
- On-Device AI Engine. Prompts Chrome's built-in Gemini Nano model via the experimental Prompt API for categorizations, falling back gracefully to local rule-based tagging if the model isn't downloaded.
What I learned
Handling dynamic social feeds means dealing with frequent class name obfuscation (such as on LinkedIn or X/Twitter). Relying on traditional CSS selectors fails immediately. Switching to stable React internally-exposed attributes (like data-urn values) proved to be the only reliable way to maintain long-term scraper resilience.
Additionally, updating complex state in a narrow browser sidepanel during fast scrolling causes DOM thrashing. List windowing and IntersectionObservers are not just optimizations — they are architectural requirements to keep chromium extension processes from crashing.
High-calibre architecture decisions.
Technical Highlights
- Manifest V3 Chromium extension with background service-worker lifecycle control.
- On-device AI integration via Chromium Prompt API and Gemini Nano optimization flags.
- IntersectionObserver-based list virtualization to prevent DOM memory overflow.
- Offline HTML generation and responsive templates for self-contained data export.
By the numbers.
If we worked together.
I bring high-leverage product engineering and absolute operational discipline: shipping weekly, pressure-testing user inputs, and automating delivery pipelines.
See it in the wild.
This case study covers how ScrollCollector was built — these links open what ships today: the live product, repo, demos, and supporting material.
The changelog.
VisaPros
Submitted as the capstone for the Google × Kaggle 5-Day AI Agents Intensive (2026). VisaPros is an agentic visa eligibility advisor utilizing a Google ADK agent-to-agent (A2A) hub-and-spoke architecture — parsing documents once and running six parallel country agents as FastAPI microservices to evaluate fit across destinations.