All work
2026Developer Platform · Security2026 — ongoing · solo build

Visitor identity & fraud prevention in one API

T screenshot
Role

Founder & CEO

Timeline

2026 — ongoing · solo build

Stack

Next.js · Go · PostgreSQL

Status

Shipped · in production

01 — Overview

What it is

TrueID is a developer platform for visitor identity, device intelligence, and fraud prevention. It gives products persistent visitor identification, rich device signals, and real-time risk scoring through a single call. I founded it and build it solo, shipping improvements every day.

02 — Problem

Why it was hard

Fraud prevention is usually a maze — separate vendors for identity, device data, and risk, each with its own integration. Smaller teams can't afford that complexity, so abusers win. The goal was one API that answers 'can I trust this visitor?' instantly.

03 — Research

What we learned first

Running operations at Orihost gave me a front-row seat to real abuse — the exact patterns fraudsters use to exploit free tiers. TrueID grew directly out of that: I could test detection against live abuse and tune it against traffic that was actively trying to get past me.

04 — Architecture

How it's built

A Go core handles device intelligence and scoring under load, with PostgreSQL for durable identity and Redis for hot, low-latency lookups. A Next.js dashboard and clean API surface make integration a single call. Risk is computed in real time on every request, not batched after the fact.

Next.jsGoPostgreSQLRedisTypeScriptDocker
05 — Gallery
trueid.me
trueid.me
trueid.me
trueid.me
06 — Features

What shipped

  • Persistent visitor identification across sessions
  • Device intelligence and fingerprinting signals
  • Real-time risk analysis on every request
  • One simple API for the whole flow
  • Battle-tested against live abuse from Orihost
07 — Challenges

Where it fought back

01

Accuracy under real abuse

Detection has to catch fraud without punishing real users. Tuning against live abuse traffic from Orihost pushed accuracy to roughly 97.98%.

02

One simple API

Hiding identity, device intelligence, and risk behind a single call — without leaking that complexity to the developer — drove the whole API design.

03

Real-time scoring

Risk answers are only useful if they arrive before you act on them. Redis-backed lookups keep scoring fast enough to gate a request inline.

08 — Results

What it did

~97.98%detection accuracy
1,000+fraudsters blocked
3+companies adopted
09 — Lessons

What I'd tell you over coffee

Build where you can see the enemy. TrueID works because it was tuned against real fraud I was already fighting, not a synthetic benchmark. Proximity to the actual problem is the unfair advantage.

Next case study

Beam