Tokenless

Tokenless

The Right Model for Each Coding Task

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
Our Smaller Model Beat Opus on Finance Tasks - Alexander Shevchenko, Ramp

Our Smaller Model Beat Opus on Finance Tasks - Alexander Shevchenko, Ramp

Alexander Shevchenko of Ramp Labs explains why an application company keeps running applied research rather than waiting for the next frontier model, and walks through three projects in production: a post-trained Qwen 3.5 specialist for financial question answering at Haiku-level latency, KV cache cartridges that prepare agents with far more context ahead of time, and a 10,000-agent swarm that pen-tested Ramp and found seven high-severity vulnerabilities using cheaper open-weight models.

AI Labs Will Trap You or Replace You - Dillon Rolnick, Nous Research

AI Labs Will Trap You or Replace You - Dillon Rolnick, Nous Research

Nous Research CEO Dillon Rolnick argues that as AI models commoditize, frontier labs will try to recoup their enormous investments by locking businesses into their platforms or competing directly in lucrative sectors like finance. He contends that proprietary software moats are disappearing, that agents behave more like operating systems than chatbots, and that companies should build on open foundations they own, such as Hermes Agent, so they can swap models and inherit new capabilities.

Robots Will Be Trained on Video Games - Pim de Witte, General Intuition

Robots Will Be Trained on Video Games - Pim de Witte, General Intuition

Pim de Witte of General Intuition argues that models acting in space and time should be built differently from turn-based LLMs. Their continuous frames-in, actions-out model is trained on Medal game recordings, since game controllers already drive most robots, and it can play games and control more than ten robot types. The talk covers world models, action density, applications such as accessibility and UX, the difference from JEPA, and upcoming early access and hackathons.

Technology

View All
What iPhone Packaging Reveals About the Robotics Opportunity

What iPhone Packaging Reveals About the Robotics Opportunity

Tutor Intelligence CEO Josh Gruenstein explains why robotics is a systems problem spanning AI, hardware, manufacturing, and customers, and how deploying robots into real factories and warehouses, with human tutors as backup, generates the data that makes them more capable. The discussion covers the company's own robot designs, its focus on small and mid-sized businesses, the economics of flexible robotic labor, and the implications for American manufacturing and supply chains.

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.

Recent Post

W&B Models end-to-end demo

W&B Models end-to-end demo

W&B Models is the system of record for the entire model development lifecycle. This guide explores how to monitor training, tune hyperparameters, track artifacts and lineage for reproducibility, and automate MLOps workflows like evaluation and deployment using a central platform.

Late-Stage Investing in an AI-Driven Market

Late-Stage Investing in an AI-Driven Market

David George (General Partner, a16z) discusses how AI is fundamentally reshaping late-stage private markets. He covers the unprecedented scale of infrastructure investment, evolving business models and monetization strategies, why AI-native companies are staying private longer, and how this new paradigm changes the calculus for durability, value creation, and long-term growth.

Artie: Real Time Data Streaming For The AI Age

Artie: Real Time Data Streaming For The AI Age

Jacqueline Cheong and Robin Tang, founders of real-time data streaming platform Artie, discuss their journey from identifying the critical need for fresh data at companies like OpenDoor to building a production-ready solution, acquiring their first major customer Substack via a cold email, and navigating the complex technical challenges of real-time data processing at scale.

The Semantic Layer and AI Agents // David Jayatillake // MLOps Podcast #343

The Semantic Layer and AI Agents // David Jayatillake // MLOps Podcast #343

David Jayatillake, VP of AI at Cube.dev, discusses the critical role of a headless, open-source semantic layer in the modern data stack. He argues against proprietary, BI-tool-specific semantic layers that create vendor lock-in and advocates for a decoupled approach. The conversation explores how AI agents can automate the entire data pipeline—from ingestion and transformation to generating and querying the semantic layer—and compares the functionalities of semantic layers and feature stores, highlighting the crucial difference of temporality.

Building Planetary-Scale Data Systems with Venice • Felix GV & Olimpiu Pop • GOTO 2026

Building Planetary-Scale Data Systems with Venice • Felix GV & Olimpiu Pop • GOTO 2026

Félix GV, an architect of LinkedIn's Venice database, discusses its unbundled, planetary-scale architecture. He covers how components like Kafka and RocksDB form independent distributed systems, details their rigorous chaos engineering practices, explains CAP theorem trade-offs in multi-region deployments, and explores the experimental integration of DuckDB for SQL-based analytics.

If You Can't See Inside, How Do You Know It's THINKING? [Dr. Jeff Beck]

If You Can't See Inside, How Do You Know It's THINKING? [Dr. Jeff Beck]

Dr. Jeff Beck explores the philosophical and technical definitions of agency, arguing that the distinction between an agent and an object lies in computational sophistication, particularly the capacity for planning and counterfactual reasoning. The conversation provides a deep dive into Energy-Based Models (EBMs), Yann LeCun's JEPA for learning in latent space, and a pragmatic approach to AI safety centered on inverse reinforcement learning rather than fears of rogue superintelligence.

← PreviousNext →

Stay In The Loop! Subscribe to Our Newsletter.

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