Background Agent Optimizations
A technical deep dive into inference optimizations for background coding agents, from semantic caching to speculative execution and smarter scheduling.
Topic Guide · 8 essays
AI Systems
Inference, agents, routing, evaluation, and model tooling.
These essays are about the machinery underneath modern AI products: inference constraints, orchestration layers, model tooling, and the developer workflows that emerge around them.
If you care more about how systems actually behave than about benchmark theater, start here. The through line is that once models are useful, the hard problems move into routing, evaluation, interfaces, and operations.
The Clearinghouse for Code
AI code generation is flooding repos with changes. The next bottleneck in software is clearing, sequencing, testing, and deploying them safely.
Teaching an LLM to Read a Profiler
AI models, reasoning, tool use, and real world data.
A Bitter Lesson
Building basketball shot-classification models taught me a classic ML lesson: sometimes human operations beat a more complicated AI system.
Vector DBs
Exploring the pivotal role of vector databases in managing multimodal data, including video and voice content.
Prototyping ML Models
Explore how Attentive leveraged LLMs to rapidly prototype a product recommendations model, significantly reducing development time from months to just a week.
LLMs in ML
Explore the integration of LLMs in enhancing traditional ML workflows at Attentive, focusing on data cleaning and organization.
Next Top Model
Discussing an innovative ML Recommendation Engine, this blog post explores a self-serve portal idea for companies to match their data with suitable ML…