Tokenless

Tokenless

AI-Friendly Code: The Missing Link for Speed with Quality • Adam Tornhill • GOTO Copenhagen 2026

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
The New AI Agents That Shop, Plan, and Gossip

The New AI Agents That Shop, Plan, and Gossip

David Pawlan, creator of Assistant Benchmark, joins Anish Acharya and Josh Elman to take stock of an agent landscape that changes almost weekly. Drawing on tests of more than 140 assistants and a survey of AI power users, they explore why trust is the dominant fear, how fences and permissions might make agents safe, and where privacy, embarrassment, and judgment fit in. The conversation then moves to multiplayer and gossipy agents, agent-to-agent networks that could spark serendipity, and the possibility that the most valuable agents will do more than save time, becoming a kind of secret superpower.

AI-Friendly Code: The Missing Link for Speed with Quality • Adam Tornhill • GOTO Copenhagen 2026

AI-Friendly Code: The Missing Link for Speed with Quality • Adam Tornhill • GOTO Copenhagen 2026

Adam Tornhill explores why AI agents produce fast early gains that often erode under growing code complexity, and how the Code Health metric, safeguarding loops, and codebase-specific refactoring can keep agentic development sustainable at scale, including a 300,000-line legacy case study.

Don't Approve the AI Model. Approve the App - Doneyli De Jesus, Langfuse

Don't Approve the AI Model. Approve the App - Doneyli De Jesus, Langfuse

Doneyli De Jesus, AI engineering lead at Langfuse, describes how a top-five bank built deployment gates for AI agents in an on-premises, air-gapped environment. He covers golden data sets, multi-dimensional scoring, why a model that was 97% grounded still failed on numbers until it was given a calculator tool, and how domain experts set thresholds and review failures, cutting evaluation time from weeks to minutes.

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

How A Team Of 7 Keeps Breaking AI Benchmark Records

How A Team Of 7 Keeps Breaking AI Benchmark Records

Poetiq, a startup by former DeepMind researchers, has developed a recursive self-improvement meta-system that builds "reasoning harnesses" on top of existing LLMs. This approach avoids the costly "fine-tuning trap" and has achieved state-of-the-art results on benchmarks like ARC-AGI and Humanity's Last Exam by automatically optimizing prompts and discovering novel reasoning strategies.

A Playground for AI Engineers

A Playground for AI Engineers

Paulo Vasconcellos from Hotmart details their journey of building "Agent as a Product", explaining how they blend classic ML models with LLMs for efficiency, evolve their MLOps platform for the generative AI era, and create real business value through AI-powered tutors and sales agents.

SW Design, Architecture & Clarity at Scale • Sam Newman, Jacqui Read & Simon Rohrer

SW Design, Architecture & Clarity at Scale • Sam Newman, Jacqui Read & Simon Rohrer

Experts Sam Newman, Jacqui Read, and Simon Rohrer explore the nuances of software design, its intersection with architecture, and the critical role of communication in scaling technical clarity. The discussion covers practical advice on implementing Architectural Decision Records (ADRs), the evolving role of the architect as a facilitator, and strategies for creating agile enterprise architectures.

Mainframe modernization explained: COBOL and AI

Mainframe modernization explained: COBOL and AI

Experts from IBM discuss the nuanced role of AI in mainframe modernization, the immense infrastructural and product challenges behind global AI adoption, and the critical need for a multi-layered, security-by-design framework for the safe deployment of AI agents.

Dylan Patel Explains the AI War While Cooking | In-Context Cooking

Dylan Patel Explains the AI War While Cooking | In-Context Cooking

Dylan Patel of SemiAnalysis discusses the AI arms race, highlighting the massive $200B+ hyperscaler capex, the true semiconductor bottlenecks shifting from data centers back to fabs, the geopolitical chess game surrounding Taiwan and TSMC, and Nvidia's strategic battle against vertical integration.

Hardware Realization and Implementation Security Evaluation of HQC, A NIST PQC Standard

Hardware Realization and Implementation Security Evaluation of HQC, A NIST PQC Standard

This talk by Sanjay Deshpande from Northwestern University explores the critical transition to Post-Quantum Cryptography (PQC) in response to the threat quantum computers pose to current public-key algorithms. It provides a deep dive into the Hamming Quasi-Cyclic (HQC) algorithm, a code-based candidate for NIST standardization. The session focuses on the challenges and innovations in creating efficient and secure hardware implementations of HQC, covering performance optimization for polynomial multiplication and countermeasures against side-channel attacks.

← PreviousNext →

Stay In The Loop! Subscribe to Our Newsletter.

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