The Problem: Sales Calls Without Convincing Demo Data
At Clevergy, we faced a persistent pain point: our sales colleagues didn't have adequate, realistic demo data to present and sell our platform to potential clients.
In the B2B energy and software space, showcasing empty dashboards or flat sample numbers ruins the pitch's credibility. While various solutions were evaluated internally, none ever saw the light of day. The product roadmap consumed all engineering bandwidth: while universally acknowledged as important, demo environments were never deemed top priority.
Faced with this bottleneck, I decided to take initiative and tackle it myself.
The First Attempt and the Rise of AI Tools
Right around that time, early generative AI dev tools like Lovable and Bolt were gaining serious momentum. Intrigued by what could be accomplished quickly with LLMs, I jumped in to build a solution.
My initial idea was to generate synthetic demo data:
- I attempted to populate the database with mock accounts, simulated telemetry, and fake consumption patterns.
- Why it failed: Energy consumption curves and invoices are highly specific. The synthetic graphs simply didn't look authentic, and we couldn't easily generate the required data volume. Moreover, keeping mock data synchronized and refreshed daily turned into an unmanageable logistical mess.
That led me to the breakthrough question: What if instead of inventing fake data from scratch, we use real client data and simply mask all sensitive identifiers on screen?
The Birth of Blurgy: Real-Time Anonymization
That paradigm shift gave birth to Blurgy. The premise was straightforward: a Chrome extension capable of blurring, masking, or replacing company names, tax IDs, addresses, tariffs, and confidential numbers on the fly while sales reps shared their screen.
I dedicated evenings, weekends, and spare moments to craft it:
- Iterating with AI Prompts & Code: Leveraging Cursor and LLM assistants to handle DOM script injection, mutation observers, and CSS element targeting.
- Brand Identity & Assets: Designed the visual branding, custom iconography, and our mascot (a cute hedgehog) alongside clean UI menus.
- Showcase Landing Page: Launched an official landing page at alejandroramirezj.github.io/Blurgy so anyone could test the extension and understand how it works.
The extension went beyond simple blur effects: it offered instant keyboard shortcuts for live calls, smart placeholder replacements to maintain layout integrity, and customizable element targeting.
The Ultimate Payoff: From Extension to Production Code
What makes me proudest about this journey isn't just shipping the extension—it's how it fundamentally changed the actual product.
Almost the entire sales team quickly adopted and relied on Blurgy for daily customer demos. However, because it operated as an external extension injecting heavy DOM mutations and styles over dynamic single-page applications:
- It experienced slight rendering delays when applying blur to dynamically loaded views.
- The sheer amount of injected code occasionally stuttered the user experience.
Rather than accepting these trade-offs, I sat down with Clevergy's Lead Frontend Engineer. I demonstrated how vital the extension had become for sales deals and explained the technical performance hurdles we were hitting on client machines.
Together, we executed a complete architectural rethink:
- Sensitive Fields Audit: We mapped out every single confidential data point across all views in the platform.
- Introducing the "blurg" Variable: Inspired directly by the project's name (a blend of Clevergy and Blurgy), our lead frontend engineer integrated a native property and directive called
"blurg"directly into the frontend codebase. - Operations Portal Command: A native toggle was added to the internal Operations Portal where sales presentations are configured. With a single command, the application activates demo mode and effortlessly obfuscates all
blurg-tagged elements—zero latency, no external extensions, and pixel-perfect aesthetics.
Furthermore, this practice became a permanent team standard: every time new features or views are built, engineers automatically tag sensitive fields with the "blurg" attribute.
Conclusion: Solving Real Problems by Closing the Loop
Today, our sales team relies on this native capability daily in virtually every prospective client presentation.
Blurgy encapsulates how I approach product innovation:
- Identify a real operational friction that the core roadmap cannot immediately prioritize.
- Harness emerging AI tools to prototype quickly and prove tangible value in days, not months.
- Collaborate cross-functionally with engineering to elevate a scrappy hack into an enduring, high-impact product feature.
Project Links
- 🌐 Blurgy Website & Demo: alejandroramirezj.github.io/Blurgy