Systems Builder & Operator

I build integrated systems —
and run the whole operation on them.

AI-powered automation · content → funnel → CRM → command center

Self-taught, hands-on. My eye for what's off came from the work — reviewing 200+ business sites a day at Google, quality-checking sensitive data at Informdata. Building my own tools powered that eye up: now I don't just spot the gap, I build the system that closes it — and operate it day to day.

30+ Python tools shipped 1 integrated stack, run mostly solo command center ↔ CRM ↔ chatbot 6+ yrs quality-grade work
See the system ↓ View the live build →
Ronald Adrian Sarinas
Who I am

A builder, shaped by quality work

I'm a self-taught systems builder. I don't build for its own sake — I build to make an operation run itself, then I run it. The thing I keep coming back to: building my own tools made my day-to-day seamless, where everything talks to everything and the busywork disappears.

The eye for detail is real, and it's earned — years on Google's team reviewing hundreds of business websites a day for what worked and what didn't, and nights at Informdata quality-checking sensitive records under strict standards. When you've inspected that much, that carefully, you notice what's off fast.

Then I started building. The eye got powered up by tool-building — when you build the pipes, you feel the moment something leaks, and you're the one who can fix it.

The short version

2023 → nowInformdata — data QA + internal automation (Power Automate, Copilot agents, dashboards)
2021 → nowDuberyMNL — founder & builder of a full AI automation stack
2019 – 23Google / TDCX — reviewed 200+ business sites/day at Google-grade standards
The build

One integrated system, run mostly solo

DuberyMNL isn't a landing page — it's an operation. I built each piece and the wiring between them: AI content & image generation, a storefront and funnel, a Messenger chatbot that captures leads into a CRM, Meta ad operations, and a command center that ties it all into one place. The payoff isn't a vanity metric — it's that what would normally take a small team, one person can run.

🎨

Content & Image Pipeline

AI captions + product/lifestyle visuals, generated on-brand at scale

→
🛒

Storefront & Funnel

Landing page + product pages, the path from click to order

→
💬

Messenger Chatbot

Answers, qualifies, and closes — with guardrails

→
🗂️

CRM / Lead Capture

Every real lead captured, scored, and followed up

🛰️

Command Center

Monitors & controls all of it — orders, inventory, KPIs, lead recovery, bot controls. Zero-terminal operation.

↑ feeding the top of the funnel: Meta Ads operations — campaign control, insights, and daily scheduling via the Marketing API + GitHub Actions.
command-center · localhost
Operations · Today
● live
18
Orders this week
5
To deliver
3
Low-stock SKUs
7
Leads to follow up
📦 Inventory tracking
Bandits Matte Black42
Bandits Tortoise27
Outback Green8
Rasta Brown3
📅 Content & scheduling
Today 6:00 PM — FB feed postqueued
Story rotation (every 3h)auto
Tomorrow — UGC reeldrafted
Calendar moment: weekend pushplanned
🗂️ CRM · lead pipeline
4HOT
11WARM
23COLD
High-intent leads to recover2 to review
🚚 Orders & deliveries
Order #1042packing
Order #1041out for delivery
Order #1039delivered
Order #1038delivered
Representative view of the command center I built — content, scheduling, inventory, deliveries, and CRM in one place. (Illustrative mock; the live app runs the real data.)
What I build with

The builder's toolkit

I work across the stack — automation, AI generation, integrations, and the dashboards that make it all observable.

⚙️

Automation & pipelines

The plumbing that removes the busywork.
Python (30+ scripts)GitHub Actions (cron / CI)Power AutomateOffice ScriptsWebhooks & APIs
🤖

AI content & generation

On-brand visuals and copy, at scale.
Nano Banana 2 / GeminiVertex AIkie.aiPrompt engineeringImage/content pipeline
🔗

Ops & integration

Making separate tools behave like one system.
Meta Marketing APIMessenger chatbot (Flask + Cloudflare Workers)Google Sheets / DriveCloudflare Pages
📊

Data & dashboards

If I can't see it, I can't run it.
Command center (live ops)Power BIChart.jsReconciliation pipelinesInventory / orders
The eye

I catch what's off — because I built it

The detail-eye is a real skill, earned from quality-grade work and sharpened by building. When you own the system, you don't just file the bug — you fix it.

The instinct is earned — years reviewing sites at Google, checking data at Informdata. The upgrade is that I now pair it with the data: I don't guess where users struggle, I read it.

Watching the live site's behavior in Microsoft Clarity, I caught what you'd never see by eyeballing a design — dead clicks (people tapping things that looked clickable but didn't respond), and a slow load I traced to heavy, uncompressed images. I compressed them, and the page started painting fast.

Each signal became a fix on the list. That's the loop I run: build it → watch how real people use it → catch what's off → fix it.

I use Clarity and the Meta pixel as signal to guide the build — practical, not a specialty I'd oversell.

Behavior insightsMicrosoft Clarity · 224 sessions
38
Dead clicks
13%
Quick-backs
6
Script errors
0
Rage clicks
Flagged → fix list
🖱️
Dead clicks detected

Users tapping elements that don't respond — 38 across the funnel

flagged
⚡
Slow load → heavy images

Uncompressed images bloating load — compressed them, faster paint

fixed
↩️
Elevated quick-backs

Visitors leaving fast after landing — friction signal

investigating