K.N.

What I've done.

I work the product side, building things with technology that real people actually use, engineering solutions to problems worth solving. Sometimes that's something tiny I just wanted to exist. Sometimes it's big enough that the challenge is the whole point. What follows are the parts that became real, the roles I've held and the things I owned and shipped.

Experience
Founding Engineer
ThriviaSep 2025 - Present

Joined pre-revenue, pre-launch as the first engineer. Took the product from pre-launch to product-market fit in under six months.

Founder
AttendanceLabs.comJan 2025 - Jul 2025

A facial recognition attendance platform for classrooms. Took the MVP from idea to launch, shipped it, ran it, then made the call to shut it down.

How I work

Started with Python, dabbled in Rust, moved through TypeScript and React, picked up Docker and deployments along the way. Now I'm stack-independent. Give me a problem and I'll figure out the tools. Power user of Claude Code and AI tooling.

Lately the problems I care about have moved past the code. I think as much about distribution and user behavior as I do about the stack, why people adopt a thing, what makes it spread, how a product actually finds the people it's for. I've run GTM and watched what converts, and learned that shipping is only half the work. The engineering still matters. It's just become the way I get at the bigger problem, not the whole of it.

For the Record

What I built.

A record of what I've actually built and shipped. I'm keeping it here as the proof behind the values, even as the rest of the site leans more into ideas and experience.

01

ATTENDANCELABS

Facial Recognition Attendance Platform.

Facial recognition attendance system for classrooms. Built with Flask/Django microservices, InsightFace ML models, and Qdrant vector database. Took it from idea to live product.

Next.jsFlaskInsightFaceQdrant
Impact
Attendance in 1 tap
ATTENDANCELABS_v1.0● SCANNING
ANALYZING...
CAPTURE
DETECT
EMBED
MATCH
RECORD
CONFIDENCE
---
EMBEDDING
512-dim
QDRANT
SEARCHING
02

BINKEDIN

Full-Stack Social Platform.

LinkedIn-inspired networking platform with Rust backend and Next.js frontend. Features authenticated login, posts, comments, likes, and follower system with PostgreSQL.

RustAxumNext.jsPostgreSQL
API Routes
6 Modules
BINKEDIN_APIPORT: 3000
U1U2MEU3U4PENDING
ACTIVITY
📝U1POST
U2LIKE
U3FOLLOW
💬U4COMMENT
POSTS
247
USERS
1.2k
FOLLOWS
3.4k
LIKES
12k
03

REQUEST LOGGER

Django Middleware for Analytics.

Plug-and-play middleware that visualizes request statistics, error rates, and endpoint performance. Tracks traffic patterns and identifies bottlenecks with time-series charts.

PythonDjangoMiddleware
Setup Time
<5 min
DJANGO_REQUEST_LOGGER● LOGGING
REQUESTS
1,247
AVG TIME
18ms
2XX
94.2%
4XX
5.8%
04

NOCTURNAL

Discord Community Bot.

Discord bot with 40+ commands for moderation, games, and community engagement. Features PostgreSQL state management, API integrations, and real-time WebSocket events.

PythonDiscord.pyPostgreSQLWebSockets
Commands
40+ Built
N
Nocturnal
● Online
discord.py
COMMANDS 40+
SERVERS 12
PING 42ms
05

MERN NINJA

Claude Code Plugin for MERN Stack.

AI-powered code reviewer and test generator for MongoDB, Express, React, and Node.js projects. Automated security audits, test gap analysis, and 1500+ lines of curated best practices.

Claude CodeMERNJestVitest
Coverage
1.5k+ Docs
MERN_NINJA● SCANNING
1const app = express();
2app.use(cors());
3app.get('/users', async (req) => {
4 const id = req.query.id;
5 const user = await db.raw(id);
6 res.json(user);
7});
ISSUES FOUND
LINES 0/7
ISSUES 0
TESTS GENERATING

More experiments and open source contributions on GitHub