John

Olumide

I architect ML pipelines, wrangle LLMs into production, and build data systems that actually scale. Currently building matchloom.ai

Scroll

Machine Learning◆LLMs◆Data Engineering◆AI Agents◆Deep Learning◆MLOps◆Data Pipelines◆Python◆TensorFlow◆Machine Learning◆LLMs◆Data Engineering◆AI Agents◆Deep Learning◆MLOps◆Data Pipelines◆Python◆TensorFlow◆

About

john@macbook — zsh

~ whoami --json

Local Time--:--:--Loading...

Building Now

matchloom.ai

AI-powered matching platform — semantic profile analysis, intelligent recommendations, and real-time matching.

matchloom.ai/Jon

Philosophy

if data == messy:

clean_it()

if model == slow:

optimize_it()

if pipeline == broken:

fix_at_3am()

# ship it. always.

AI & DataWhere I live and breathe

BuilderShip > Perfect

FastPrototype to prod, rapidly

Tech Stack

AI & Machine Learning

TensorFlowPyTorchLangChainLangGraphLLMsDeep LearningNLPHugging FaceScikit-learnMLflow

Data Engineering

SnowflakeDBTAirflowApache SparkKafkaSQLPostgreSQL

Backend & Infrastructure

PythonFastAPIDockerAWSCI/CDREST APIsRedisTerraform

What I'm Building

Flagship Project

matchloom.ai

AI-powered intelligent matching platform. Semantic profile analysis using LLMs, real-time recommendations, async Celery workflows, Redis-cached matching engine, and a FastAPI backend deployed on AWS ECS Fargate.

matchloom.ai

ML / Deep Learning