The Challenge: Heavy Payloads, Unpredictable Compute, and Thousands of Models To an external client, our serverless inference API looks like a standard web service: you POST an image and get back a clean JSON block of predictions, usually in under 100 milliseconds. Behind that simple promise sits some genuinely gnarly engineering. If your background is mostly in…
16th July 2026 – Link Blog Firefox in WebAssembly (via) This is absurdly cool: Puter compiled Firefox to WebAssembly such that the whole browser runs in another browser. Here’s my blog, running in Firefox, running in WebAssembly, running in Chrome: They chose Firefox/Gecko because it has strong single-process support. The project used an estimated $25,000…
In Part 2, we built the three pillars of Structured Mojo Kernels: TileIO for data movement, TilePipeline for coordination, and TileOp for compute. These three pillars form a strong structured abstraction of kernel programming – enabling composition of modular typed and testable pieces rather than a large monolithic block of interleaved logic. This post shows…
What You Should Know Healthcare AI infrastructure company Bunkerhill Health has closed its Series B funding round led by Khosla Ventures, bringing its total funding to $55M across its Seed, Series A, and Series B rounds. The company’s core platform, Carebricks, empowers health systems to build and deploy their own AI agents across clinical, operational,…
talk about context engineering today, they usually mean retrieving the right context from a document: chunk selection, hybrid search, reranking, TOC-aware retrieval. That is one half of the story. The other half is that the question itself is context the LLM will see, and the question deserves the same treatment as the retrieved passage. Take…
Weather predictions need to capture the full range of possibilities — including worst case scenarios, which are the most important to plan for. WeatherNext 2 can predict hundreds of possible weather outcomes from a single starting point. Each prediction takes less than a minute on a single TPU; it would take hours on a supercomputer…
For most large enterprises, AI transformation is a story of silos — business units running competing experiments, data sitting in disconnected systems, and outcomes staying out of reach. Cushman & Wakefield has taken a different path. Sal Companieh, Chief Digital and Information Officer for the global commercial real estate services firm, has spent the past…
Mira Murati’s Thinking Machines Lab Inc. today launched its first foundation model with the release of Inkling, making its full open weights available to developers so they can fine-tune it as they wish. Inkling is the first model fully trained from scratch by Thinking Machines, coming after a year in which the company mostly made…
Retrieval is critical in multi-step agentic workflows where poor retrieval can cause agents to fetch irrelevant context, re-query, waste token budget, and carry noise into later reasoning steps. Today, we are releasing NVIDIA Nemotron 3 Embed, a collection of open and commercially available embedding models designed to improve retrieval quality while giving developers practical deployment…
Capcom’s RE ENGINE team set out to bring path tracing into two shipping titles at once, Resident Evil Requiem and PRAGMATA, each with a different visual identity. Over two years, the RE ENGINE team built a reference path tracer, validated its lighting against common DCC path tracers, then scaled it into a game-oriented version optimized…