Horizon LabsHorizon Labs

Insights

Product, design, AI, and engineering perspectives from our team.

Feature Stores for Machine Learning at Scale
Data foundations27 Aug 2026

Feature Stores for Machine Learning at Scale

Growing data teams often outgrow spreadsheets and ad hoc pipelines long before they need a full feature store platform. Here's how to tell the difference — and how Google Cloud's Vertex AI implements feature management as a concrete example.

7 min readChris Kerr
ML Proof of Concept to Production: Why Most Stall
AI in production1 Aug 2026

ML Proof of Concept to Production: Why Most Stall

Most ML proof of concepts never reach production — not because the model is wrong, but because reproducibility, deployment, monitoring, evaluation, versioning, and lineage were never built. Here's why POCs stall and what a realistic path to production looks like.

6 min readChris Kerr
MLOps Consulting in Australia: From Notebook to Production
Choosing a partner27 June 2026

MLOps Consulting in Australia: From Notebook to Production

MLOps consulting helps Australian engineering teams close the gap between a model that works in a notebook and one that reliably runs in production. This guide covers the MLOps maturity model, five core capabilities, tooling options including MLflow, Kubeflow, SageMaker, and Vertex AI, and the Australian data residency and privacy obligations that affect how ML pipelines should be architected.

9 min readChris Kerr
PyTorch vs TensorFlow in 2025: Choosing a Production ML Framework
AI in production7 June 2026

PyTorch vs TensorFlow in 2025: Choosing a Production ML Framework

PyTorch and TensorFlow are both production-capable ML frameworks in 2025, but they suit different teams, workloads, and deployment environments. This guide helps technical leaders make a defensible framework choice based on ecosystem fit, serving requirements, and team context — not benchmarks or hype.

11 min readChris Kerr
Data Quality for AI: Why Garbage In Still Means Garbage Out
Data foundations12 Apr 2026

Data Quality for AI: Why Garbage In Still Means Garbage Out

Poor data quality is the fastest way to turn a promising AI project into an expensive failure. Learn how to assess if your data is AI-ready and implement a practical framework for data quality: profiling, validation, monitoring, and remediation.

7 min readJames Liu
MLOps Explained: How Production AI Stays Reliable After Launch
AI in production29 Mar 2026

MLOps Explained: How Production AI Stays Reliable After Launch

MLOps ensures AI models remain accurate and reliable in production through continuous monitoring, automated retraining, and governance frameworks. Learn how to detect model drift, implement monitoring pipelines, and build automated retraining systems that keep production AI performing at peak effectiveness.

7 min readSarah Mitchell
Predictive Maintenance with Machine Learning: Implementation Guide
AI strategy & ROI29 Mar 2026

Predictive Maintenance with Machine Learning: Implementation Guide

Learn how to implement predictive maintenance with machine learning, from sensor data pipelines to model deployment. Includes a detailed case study showing 84% downtime reduction in Australian mining operations.

12 min readJames Liu