BUC Populi - AI Grading & Compliance, Automated
A weekly cron pipeline that grades, runs an attendance-compliance check, and emails professors - with a human approving every grade.
M.Sc. Computer Science·Westcliff University Irvine·MBA·Bethesda University of California·M.Sc. E-Business·SGH Warsaw School of Economics
Biography
I find the work people do by hand-the repetitive, error-prone, easy-to-forget tasks-and put reliable automation behind it. AI does the first pass at scale, scheduled pipelines keep the cadence, and a human stays on the decisions that carry real stakes. The goal is never "let the machine take over"; it's to delete the toil and raise the quality of what's left. I've built these workflows in production-grading and compliance pipelines, scan-to-value flows, trial-to-paid funnels, and content drops-where automation is the product, not a side script.
Skills
Years Experience
10+
Projects
28
Manual Hours Removed
60%+
Services
An engineer who turns manual workflows into dependable automation-AI for the first pass, pipelines for the cadence, and a human on every decision that matters.
Map the manual process end-to-end, then replace it with a scheduled pipeline-fetch, transform, act, notify-that runs on its own and never forgets a step.
Put an LLM on the first pass-drafting, classifying, flagging at scale-while a human stays the system of record on anything with real stakes.
Instrument what the automation does-time saved, errors caught, exceptions flagged-so the workflow demonstrably raises quality, not just throughput.
Selected Projects
Production workflows where manual toil became reliable, measurable automation
A weekly cron pipeline that grades, runs an attendance-compliance check, and emails professors - with a human approving every grade.
Phone-camera card identification and daily-recomputed pricing automating a process that used to take minutes per card, in an app with 85K+ accounts.
An automated trial-to-paid funnel-onboarding, entitlements, and server-verified subscriptions-running without anyone in the loop.
One source of truth for character data that ships the playable roster and the marketing drops from the same pipeline-no manual re-entry.
Writing
Notes on engineering, AI, and building products that scale
Not every manual task deserves automation, and not every one is safe to hand to a machine. Before you write a single line, run the work through Problem, Manual, Solution, Automation.
The point of AI in a workflow isn't to replace the person-it's to delete the repetitive work so the person can spend their attention where it actually matters.
A button that does one step faster is still a manual workflow. Real efficiency comes from automating the whole sequence-fetch, transform, act, notify-so nobody has to remember it runs.
How it starts
A 15-minute call
After that
A reply within a day
Based in
Long Beach, CA · remote friendly