Tokenless

Tokenless

Why We Rebuilt Microsoft Office From Scratch for AI Agents - Yijia Liang, Compound AI

Interactive discovery

Explore the topic map

Follow the connections between themes, people, and ideas across the Tokenless archive in an interactive topic modeling map.

Open the map Browse posts

Machine Learning

View All
Judge a Transaction by the Company It Keeps — Somya Rai & MD Sharique, EXL

Judge a Transaction by the Company It Keeps — Somya Rai & MD Sharique, EXL

EXL developed an advanced fraud detection system that augments XGBoost with Graph Neural Network (GNN) embeddings. Traditional tabular models struggle to capture fraud signals residing in transaction neighborhoods. The system's innovative design includes meticulous sampling strategies, disjoint monthly graph construction for leakage control, and a GraphSAGE architecture that aggregates mean, max, and standard deviation of neighbor features to detect subtle fraud patterns. Emphasizing explainability for regulatory compliance and analyst trust, it also features a highly optimized serving architecture achieving 3ms P99 latency, pivoting to edge prediction to reduce cache footprint from 60GB to 2-3GB. This led to a 27-33% relative lift in true positive rates and a projected $65 million incremental fraud capture.

From 15% to 90% GPU Utilization: Fix the Data Pipeline, Not the Model

From 15% to 90% GPU Utilization: Fix the Data Pipeline, Not the Model

Tarun Sunkaraneni from Amazon AGI shows how a multimodal SFT pipeline that spent about 85% of its time waiting on data was brought to roughly 90% GPU utilization. He walks through concurrency with asyncio and Ray actors, prefetching, zero-copy transfer via Ray's object store, and a network-card bottleneck that appears only at scale, and argues that the hardest part of training throughput is often keeping GPUs fed rather than optimizing the model.

Small Models, Big Results: Training a Finance Agent for Under $500 — Charles Dickens, Snorkel AI

Small Models, Big Results: Training a Finance Agent for Under $500 — Charles Dickens, Snorkel AI

Snorkel AI and UC Berkeley's Sky Computing Lab show that a 4B Qwen3 model, trained with reinforcement learning for under $500 using the rLLM framework, beats its 235B sibling on expert-validated financial question answering built from SEC 10-K filings. The talk covers the FinQA dataset construction and verification, common failure modes of frontier models, transfer to multi-table and general tool-use tasks, and ablations showing that simple data and binary rewards outperform more elaborate setups, since the real bottleneck is disciplined tool use rather than reasoning depth.

Artificial Intelligence

View All
Why We Rebuilt Microsoft Office From Scratch for AI Agents - Yijia Liang, Compound AI

Why We Rebuilt Microsoft Office From Scratch for AI Agents - Yijia Liang, Compound AI

Yijia Liang, founder of Compound and former finance analyst, explains why personal agents succeeded and what finance needs beyond them: multiplayer agents with their own inboxes, a headless Office engine rebuilt with agents, and agent swarms that navigate massive data rooms and models while keeping every output grounded in citations.

Most of Your Docs Readers Are Now AI Agents - Hahnbee Lee, Mintlify & Lincoln Murr, Coinbase

Most of Your Docs Readers Are Now AI Agents - Hahnbee Lee, Mintlify & Lincoln Murr, Coinbase

Hahnbee Lee of Mintlify and Lincoln Murr of Coinbase discuss what changes when AI agents become the main users of software products. They cover why documentation now needs to be optimized for context windows and accuracy rather than design, how Coinbase is exposing its business through APIs, and how the x402 payment standard and crypto microtransactions let agents pay for services and gain greater autonomy.

If Your AI Scores 100%, Your Eval Is Wrong - Ramakrishnan Lokanathan, eslams.com

If Your AI Scores 100%, Your Eval Is Wrong - Ramakrishnan Lokanathan, eslams.com

Ramakrishnan Lokanathan argues that AI evals are broken because they are overcomplicated, and lays out a practical framework for building them for high-stakes applications: define the most ambitious version of success, test every step of a workflow, combine automated checks with human review and adversarial tests, guard against regressions, and turn real user failures into tests. He also introduces eSlams, an open-source package where models compete in games like chess, Go, and poker as a window into how they reason.

Technology

View All
Why Founders Are Betting Big on Factories, Mines, and Energy

Why Founders Are Betting Big on Factories, Mines, and Energy

Erin Price-Wright and Connor Love unpack the AD50 and the rebuilding of America's industrial base, covering how defense demand is financing manufacturing capacity for commercial markets, how AI is reshaping factory economics, how a new generation of founders applies lessons from SpaceX, Tesla, Anduril, and Palantir to mining, energy, and manufacturing, and why private capital is now willing to fund physical infrastructure before government demand is guaranteed.

Mark Cuban’s Plan to Make Healthcare Affordable

Mark Cuban’s Plan to Make Healthcare Affordable

Mark Cuban and Julie Yoo discuss how transparency, direct contracting, and AI could reshape healthcare economics, from Cost Plus Drugs and modular pod manufacturing to personalized medicine, the future of physicians, and advice for founders who want to use AI to learn rather than to stop thinking.

Exploits Now Beat Patches: Open Source Security in the AI Era - Patrick Smyth, Chainguard

Exploits Now Beat Patches: Open Source Security in the AI Era - Patrick Smyth, Chainguard

Patrick Smyth of Chainguard examines how frontier AI has compressed the time between vulnerability disclosure and exploitation, leaving 30-day remediation cycles obsolete. The talk covers AI-assisted attacks, waves of supply chain malware, four attack surfaces spanning malware, known CVEs, zero days, and side doors like CI/CD, developer laptops, and agent skills, and practical defenses including cooldown periods, faster CVE remediation, and signing artifacts with SLSA and Sigstore.

Recent Post

AI Agents Need Computers: 74% MoM Growth, 850K/Day Runs, & New Agent Cloud — Ivan Burazin, Daytona

AI Agents Need Computers: 74% MoM Growth, 850K/Day Runs, & New Agent Cloud — Ivan Burazin, Daytona

Daytona CEO Ivan Burazin discusses the company's pivot from developer environments to composable computers for AI agents. He explains the unique infrastructure challenges posed by spiky RL and eval workloads, Daytona's bare-metal architecture with a custom scheduler for high performance, and the future need for stateful Windows and macOS sandboxes to automate knowledge work.

Cooking with Agents in VS Code — Liam Hampton, Microsoft

Cooking with Agents in VS Code — Liam Hampton, Microsoft

Liam Hampton from Microsoft presents a practical framework for using AI agents effectively by categorizing them into three types: local, background, and cloud. He demonstrates how to run all three simultaneously from a single VS Code interface to solve separate problems in one codebase, showcasing a powerful, integrated developer workflow.

Scaling Agents on Kubernetes with acpx and ACP — Onur Solmaz, OpenClaw

Scaling Agents on Kubernetes with acpx and ACP — Onur Solmaz, OpenClaw

Onur Solmaz from OpenClaw discusses the challenge of managing 300-500 daily, often AI-generated, pull requests. He introduces ACPX, a headless CLI for the Agent Client Protocol (ACP), designed to automate PR triage through a node-based workflow. The talk culminates in a vision for on-demand, disposable agent pods on Kubernetes, managed by a Go operator that provisions and tears down full compute environments per task, wiring them into chat platforms like Slack.

How to Build a Self-Improving Company with AI

How to Build a Self-Improving Company with AI

YC General Partner Tom Blomfield explains how to move beyond the traditional hierarchical company structure and build a self-improving organization using AI. He introduces the concept of recursive, self-improving AI loops that can optimize a company's operations, products, and knowledge base while the founders sleep.

Your Coding Agent Should Do AI System Engineering — Ben Burtenshaw, Hugging Face

Your Coding Agent Should Do AI System Engineering — Ben Burtenshaw, Hugging Face

Ben Burtenshaw from Hugging Face demonstrates how coding agents are tackling complex AI systems engineering tasks. He outlines a three-tiered approach: interactively writing CUDA kernels, autonomously fine-tuning LLMs, and deploying a multi-agent research lab (AutoLab) to parallelize experiments, all powered by file-based "skills" and open primitives on the Hugging Face Hub.

"Garbage In, Garbage Out" is a LIE

"Garbage In, Garbage Out" is a LIE

Terrence Lee-St. John, author of "From Garbage to Gold," challenges the "garbage in, garbage out" mantra. He presents a data-architectural theory explaining why models trained on noisy, high-dimensional tabular data can achieve robust predictive performance by focusing on recovering latent signals rather than exhaustive data cleaning.

← PreviousNext →

Stay In The Loop! Subscribe to Our Newsletter.

Get updates straight to your inbox. No spam, just useful content.