Pats Laurel

AI systems. Production ready.
Reliable. Scalable. Real-world tested.

Building AI systems that actually work in production

About

Building AI systems that work in the real world

/ WHO I AM

I build AI systems for real problems. Most of my work involves taking research ideas and making them work reliably at scale.

/ WHAT I DO

I focus on building reliable AI systems that don't hallucinate, and creating AI agents that can handle edge cases.

/ PHILOSOPHY

I care more about systems that work than systems that sound impressive in demos. I also enjoy sharing knowledge through talks at conferences, universities, and tech communities.

/ BASED IN

Metro Manila, Philippines

/ FOCUS

AI Engineering

/ STATUS

Available

Experience

What I do and how I approach AI engineering

/ CURRENT FOCUS

Most of my work involves taking research ideas and making them work reliably at scale. I've spent a lot of time figuring out why models break in production, how to make RAG systems that don't hallucinate, and building AI agents that can handle edge cases.

/ WHAT I BUILD

Production-ready AI systems that handle real user chaos. From RAG pipelines to AI agents, I focus on reliability over hype. Every system is designed to work at scale, not just in demos.

/ COMMUNITY

I enjoy sharing knowledge through talks at conferences, universities, and tech communities about AI engineering and production ML systems.

/ APPROACH

Ship fast, measure everything, fix what breaks. I believe in pragmatic solutions over perfect ones, and in learning from production failures.

/ AVAILABILITY

Open to collaborations and side projects

Let's connect →

Tech Stack

Tools I use daily

/ Core Languages

Python

Rust

C++

/ Performance & Numerical Computing

CUDA

Triton

JAX

ONNX Runtime

TensorRT

/ Machine Learning

PyTorch

TensorFlow

scikit-learn

Hugging Face

/ Focus Areas

NLP (LLMs)

Reinforcement Learning

Causal Inference

/ LLM Development & RAG

LangChain

PydanticAI

LlamaIndex

Pinecone

Qdrant

Weaviate

Elasticsearch

OpenSearch

Faiss

pgvector

/ Inference & Deployment

Docker

Ollama

AWS Lambda

Kubernetes

gRPC

Redis

NGINX

KServe

/ Cloud & MLOps

AWS

GCP

Amazon SageMaker

Amazon Bedrock

MLflow

Weights & Biases

DVC

Apache Airflow

Dagster

Kubeflow

Evidently AI

WhyLabs

Great Expectations

Prometheus

Grafana

OpenTelemetry

/ Databases & Storage

PostgreSQL

Supabase

Amazon S3

/ Data & Streaming

Apache Kafka

Apache Spark

DuckDB

/ Full-Stack

FastAPI

TypeScript

React

Next.js

Great AI products start here,
when ideas meet execution.

Have a technical challenge? Need guidance on AI systems? Or just want to discuss production ML? Let's talk.

/ EMAIL

dev@patslaurel.com

/ LOCATION