// portfolio-v2.0

Full-Stack Developer

I build web apps with React, Next.js, and Postgres. Ship fast, break nothing, learn from users.

TypeScriptReactNext.jsNode.jsPostgreSQLTailwind

View Projects Contact Me

const developer = {

focus: full-stack,

craft: clean-ux-ex,

learn: every-day,

};

// glow born from dim + neon

Projects

Technologies

Blog posts

Ideas

/ 01 · selected work

Featured Projects

All Projects →

★ featured

Compliance Dashboard for AI Teams

I built this dashboard for a team that ran eleven AI agents across support, sales, and scheduling. Each agent logged in its own format. Release reviews took two days of searching log files and asking developers what each tool call did. The dashboard gathers the whole fleet in a single view. It streams tool calls from every agent into one audit table with session id, tenant, tool name, and policy decision. A lineage view traces each answer back to source documents and retrieval calls. Auto-flagging marks sessions with unapproved outbound messages, raw PII in logs, and retention stamps past due. Reviewers open flagged sessions first and clear the rest in bulk. Export takes one click. It produces a dated report with call counts per agent, approval coverage, and deletion counts from the nightly job. The security lead signs releases from this report, and auditors read the same pages during reviews. Review time dropped from two days to twenty minutes. The team ships weekly now, and the dashboard catches two or three policy misses per month before customers see them. I repeat the same twenty-minute review at the start of every month.

complianceaisecuritydashboard

Concierge: AI Front Desk for Local Businesses

Concierge started with one question from a plumber in Porto: who answers the phone while I fix pipes under a sink? I built the answer as a hosted front desk for local businesses. It books appointments, answers repeat questions, and captures leads after hours. Every conversation works as a funnel. The widget greets visitors with three chips: book a visit, check a price, ask a question. Booking pulls open slots from the shop calendar and confirms by SMS. The bot pulls price answers from the owner rate card, and each answer ends with an offer to book. After-hours visitors leave a name and number. The owner opens the morning lead queue with source and intent labels on each row. Owners run the product from one screen. It shows today bookings, new leads, and answered questions with thumbs-up and thumbs-down marks from visitors. I read the negative marks every Friday and add the missing answers the same day, so the bot covers more ground each week. Four months after the first install, eleven shops paid monthly and MRR reached $2,800. Setup takes thirty minutes because I preload services, prices, and hours from the shop website and Instagram. The booking link lives in the Google profile of each shop and brings visible calls every week, so owners keep paying.

agentic-uimicro-saascrolocal-business

AI Booking Agent for Healthcare

I built this booking agent for a healthcare provider with five locations and one overloaded phone line. Receptionists answered calls between check-ins, patients waited on hold, and missed calls left slots empty. The provider processed 200+ bookings per week, so every lost call cost real revenue. I shipped two surfaces. Patients meet a chat widget on the clinic website. It answers questions about doctors, hours, and visit preparation, offers open slots from the scheduling API, and confirms bookings by SMS. Staff open a dashboard. It shows the day schedule per location, flags double bookings, and lists callbacks with phone numbers and reasons. Compliance shaped the build before a single booking flowed through it. Every tool call lands in an audit table with session id and approval reference. Phone numbers pass through a redaction pass before they reach logs. Each row carries a thirty-day retention stamp with nightly deletion. The provider passed its data review before launch with zero rework. Eight weeks after launch, phone load at reception dropped by half and slot fill rose from 78 to 94 percent. The provider now runs the agent across all five locations, and staff onboard new hires with the dashboard in one sitting.

aihealthcarecomplianceagentic-ui

/ 02 · principles

How I Work

Four habits I use when the stack changes weekly.

Embrace Uncertainty

Benchmarks shift weekly. New tools ship monthly. I run experiments instead of guessing, and I update my priors when the data says I'm wrong.

Experiment & Iterate Fast

I ship a hypothesis, measure the result, and change direction based on evidence. Mistakes cost less than debates.

Shared Ownership

Engineers and product people see the same dashboards. When a feature ships, everyone owns the outcome.

Challenge Assumptions

I ask 'why do we do it this way?' before accepting a process. The best solutions come from questioning defaults, not following playbooks.

/ 03 · about

About Me

I write TypeScript, ship features, and fix production issues. React on the frontend, Node.js on the backend, Postgres for data. I've reduced deploy times from 8 minutes to 90 seconds and cut bundle size by 40%.

active stack

TypeScriptReactNext.jsNode.jsPostgreSQLRedisDockerTailwind

// 2026 · still learning

// rate-limits: none

// coffee→code: efficient

/ 04 · journal

Latest from the Blog

All Posts →

Design for Web3: Where Trust Meets Regulation

I design DeFi and crypto interfaces where a misclick costs money: explicit confirmations, readable fees, USDT payment rails. This guide covers the trust patterns regulators also want to see.

web3securitybusiness

Sep 5, 2026

Free Traffic in 2026: What Actually Works

I get clients without ad spend: short video, niche communities, email, and pages AI search cites. Here is the exact weekly system I run.

marketingbusinessworkflow

Sep 5, 2026

Blue Ocean for a Senior Designer in 2026: Five Moves

I spent 15 years designing screens, then moved where competition cannot follow: business-model design, client-trained AI, regulated niches. These five moves built my blue ocean.

businessculturestrategy

Sep 5, 2026

/ 05 · ideas

Recent Ideas

All Ideas →

Shared ownership: wins and losses are collective

The team that shares ownership of failure is the team that moves fast without blame.

teamcultureleadership

2026-08-06

Learn faster than the tech changes

Continuous learning is the sixth pillar of AI readiness. The team that learns fastest wins.

learningaiculture

2026-08-06

Intellectual honesty is a budget line

Admitting you don't know something early costs nothing. Discovering it in production costs a lot.

teamaiculture

2026-08-06

Have a project in mind?

I take a few projects each quarter. Tell me what you ship and how you measure it.

Get in Touch