Browse by theme
AI for Math
Cases from First Proof, Momus, LeanMarathon, Danus, and Aletheia reveal the central challenges math agents face in strategy selection, dependency management, long proofs, and preserving information from failed attempts.
AI for Math · Math Agent · Multi Agent · Formal Reasoning · Proof Verification · 17 min read ·
Read EssayAgent Engineering
A practical cost-optimization guide for agents, skills, MCP, AI coding, and model APIs, covering task economics, context, caching, tools, model selection, and stop conditions.
LLM Engineering · Agent Efficiency · Context Engineering · LLM Cost · Model Routing · 13 min read ·
Read EssayAgent Engineering
A field guide to seven recurring Multi-Agent failure modes: task decomposition, hidden dependencies, duplicate work, ownership, context handoffs, correlated errors, and debate.
Multi Agent · Agent Orchestration · Agent Reliability · Context Engineering · Agent Evaluation · 14 min read ·
Read EssayAgent Engineering
An AI learning roadmap for 2026 is more than a tool list. Explore six durable skills--discovery, specification, modeling, orchestration, verification, and delivery--for turning ideas into real outcomes.
AI Learning · Agent Workflow · Education AI · Career Skills · Personal Productivity · 14 min read ·
Read EssayAgent Engineering
How well can AI solve CAPTCHAs in 2026? A practical analysis of text, image grids, sliders, GUI agents, end-to-end pass rates, and modern bot risk systems.
Agent Engineering · GUI Agent · AI Security · Computer Vision · CAPTCHA · 23 min read ·
Read EssayAgent Engineering
Explore Google ScientistOne's Chain-of-Evidence (CoE) framework for verifiable AI research, from claim classification and traceability to integrity audits.
Agent Engineering · Autonomous Research · Agent Evaluation · Verifiable AI · ScientistOne · 12 min read ·
Read EssayAgent Engineering
A critical analysis of Loop Engineering through completion criteria, proxy metrics, error amplification, context corruption, engineering productivity, and agent security.
Agent Harness · Coding Agent · Agent Reliability · Agent Loop · Loop Engineering · 15 min read ·
Read EssayAgent Engineering
A practical guide to Loopcraft: why AI agent engineering is moving from one-off prompt optimization toward event triggers, verification loops, retries, state, and continuously improving agent systems.
Agent Harness · Coding Agent · Agent Orchestration · Agent Loop · Prompt Engineering · Loopcraft · 15 min read ·
Read EssayAgent Engineering
Starting from a real customer-support ticket, this piece breaks down the five stages of Agentic RAG -- orchestration, query rewriting, parallel retrieval, sufficiency checking, and synthesis -- and covers cross-system permission boundaries plus three engineering decisions on routing, iteration depth, and latency.
Agent Engineering · Agent Orchestration · Retrieval · Enterprise AI · Agentic RAG · 15 min read ·
Read EssayAgent Engineering
A practical review of three Anthropic Claude Code production incidents, covering reasoning history, system prompts, cache invalidation, default reasoning settings, and real-world AI agent testing.
Coding Agent · Agent Harness · Agent Reliability · Context Engineering · Claude Code · 10 min read ·
Read EssayAI for Math
A breakdown of how DeepMind AlphaProof Nexus combines LLMs, Lean, AlphaProof, evolutionary search, and multi-agent orchestration into a formal proof system for mathematical research.
AI for Math · Formal Reasoning · Multi Agent · AlphaProof · Deepmind · 17 min read ·
Read EssayAgent Engineering
A practical breakdown of LangChain's Runtime for production deep agents — durable execution, layered state, human-in-the-loop, permissions, middleware, streaming, observability, and a concrete design checklist for business-grade Agent Runtimes.
Agent Harness · Agent Runtime · Agent Reliability · Agent Observability · LangChain · 28 min read ·
Read EssayAgent Engineering
A practical breakdown of Anthropic Managed Agents, covering agent harness architecture, sessions, sandboxes, credentials, context builders, traces, evals, and production AI agent runtime patterns.
Agent Harness · Agent Runtime · Agent Reliability · Anthropic · Managed Agents · 18 min read ·
Read EssayAgent Engineering
A practical deep dive into Agentic Harness Engineering, coding agent harness design, agent observability, middleware, memory, evaluation, and rollback for AI engineering teams.
Agent Harness · Coding Agent · Agent Evaluation · Agent Observability · AHE · 23 min read ·
Read EssayAgent Engineering
A practical DSPy tutorial for AI builders in North America. Learn how DSPy Signatures, Modules, and Optimizers improve prompt engineering, prompt optimization, and LLM pipeline design for production AI workflows.
LLM Engineering · Prompt Engineering · LLM Evaluation · DSPy · LLM Pipeline · 11 min read ·
Read EssayAgent Engineering
A technical breakdown of PaperBanana's Retriever, Planner, Stylist, Visualizer, and Critic agents, showing how multi-agent AI generates academic method diagrams for research writing and scientific figure workflows.
Multi Agent · Agent Orchestration · Academic Figures · Agent Workflow · PaperBanana · 9 min read ·
Read EssayAgent Engineering
A focused breakdown of how PaperBanana coordinates Retriever, Planner, Stylist, Visualizer, and Critic agents to generate academic method diagrams.
Multi Agent · Agent Orchestration · Academic Figures · Role Design · PaperBanana · 6 min read ·
Read EssayAI for Math
A technical guide to AlphaGeometry DSL covering problem format, defs.txt action definitions, geometric predicates, rules.txt reasoning, and construction workflow for geometry solvers, dataset generation, and AlphaGeometry reproduction.
AI for Math · Geometry Reasoning · Formal Reasoning · Geometry DSL · AlphaGeometry · 11 min read ·
Read EssayAI for Math
How AlphaGeometry2 combines geometry reasoning, auxiliary-line search, and symbolic deduction to solve IMO geometry problems, with engineering takeaways for math AI builders.
AI for Math · Geometry Reasoning · Formal Reasoning · AlphaGeometry · Deepmind · 11 min read ·
Read EssayAI for Math
From AlphaGo to AlphaGeometry, AlphaProof, and Aletheia, this essay unpacks the engineering value of a verifiable autonomous math research loop.
AI for Math · Autonomous Research · Formal Reasoning · Math Agent · Deepmind · 10 min read ·
Read EssayAI for Math
From executable plotting to geometric proof, a structural analysis of CodePlot-CoT's strengths and boundaries.
AI for Math · Geometry Reasoning · Visual Reasoning · Code Generation · CodePlot CoT · 9 min read ·
Read EssayAI for Math
A 2026 view of AI for math, from benchmark credibility and capability tiers to architecture-heavy solving systems and real product opportunities.
AI for Math · Formal Reasoning · Math Reasoning · AI Benchmark · 11 min read ·
Read Essay