Turning Systems Into Scale
I’ve built lightweight internal tools and proof-of-concept automations that bridge support, sales, and analytics. These aren’t production products, they’re model-esque prototypes designed to solve real problems, explore technical workflows, and push process ideas from concept into clarity.
Each project reflects my focus on structure, efficiency, and data-driven impact. The same mindset I apply in business, operations, and product design.
🧩 Guardrail — Process Governance System
Guardrail acts as a compliance firewall for AI. It proxies model calls, inspects each response against configurable policies, and takes an action—allow, redact, block, or escalate—before the content reaches end users. Every decision is versioned, reason-coded, and written to an audit-ready database with webhooks for triage. Policies are code-defined (YAML), so updates are reviewed and deployed like any other change.
⚙️ CRM Dupe Remover — Data Hygiene Script
Experimented with a Python-based matching logic to identify and merge duplicate CRM records using fuzzy matching and domain normalization.
The prototype cleaned thousands of mock entries in sandbox tests and became a reference model for what a scalable deduplication automation could look like inside Zoho CRM.
📊 Service Desk KPI Dashboard — Live Metrics Hub
Modeled a Power BI dashboard integrating ticket data to visualize SLA performance, resolution speed, and trend patterns.
Though built on static exports, it introduced real-time thinking into support operations — setting the stage for metric-driven accountability across the team.
🧠 SQL Insights — Trend Detection Sandbox
