Skip to contentNew book: Designing Human-Centered AI InterfacesGet it
Abhishek Anand

Notes from the work

Essays and field notes from 17+ years of building software and teams

Mostly what I learn by shipping: agentic AI systems, frontend architecture, performance work, and the odd lesson about leading engineers. Written for practitioners, with the trade-offs left in.

Archive

More articles

Categories

Popular tags

15 articles

AI & Agents

November 16, 2026 · 11 min read

When a Decision Model Says 90%, How Often Is It Right?

Decision models return a typed answer with a probability instead of text. I sent the same 1,200 labelled items to Gemini Decision 1, GPT-6 Luna through OpenAI’s Decisions API, and Jev 1.14, and checked how often an answer stated at 90% or more was right.

LLM Evaluation · Calibration · Classification · Jev

AI & Agents

October 28, 2026 · 9 min read

Running a Studio of AI Agents: One Orchestrator, Many Producers, a Shared Board

A Claude orchestrator, one Claude owner per project, Codex producers in tmux, an append-only message board, and cheaper models added only after a measured trial. What worked, what broke, and which parts I would reuse for software engineering.

AI Agents · Claude · Codex · Multi-Agent Systems

AI & Agents

September 20, 2026 · 16 min read

Testing Jev: An AI Model Built for Decisions

Jev returns typed decisions instead of text. I tested its calibration, speed, evidence checks, and failure cases against a Gemini Flash baseline. The useful part is how it fits into software and the questions you still have to get right.

Jev · AI Tools · LLM Evaluation · Classification

AI & Agents

September 18, 2026 · 13 min read

Getting AI Agents to Delegate Work and Review Each Other’s Code

How my model-delegation skill decides what to keep in one session, what to hand to another agent, and when to ask a different model for a review. With a real example from this website and an open-source skill.

AI Agents · Codex · Claude · Gemini

AI & Agents

September 13, 2026 · 10 min read

AI Needs a Writing Standard

How I make AI responses easier to read with writing rules in AGENTS.md and CLAUDE.md, lessons from technical writing standards, and existing rewrite skills.

AI Tools · Claude · Writing · Productivity

AI & Agents

April 25, 2026 · 8 min read

Building Agent Systems with Gemini, ADK, A2A, and A2UI

What each part of an agent system is responsible for, how they fit together, and where application code still needs to make the decisions.

Agentic AI · ADK · Gemini · A2A

AI & Agents

July 14, 2025 · 7 min read

MCP Servers: Connecting AI Assistants to Your Tools

A practical guide to MCP architecture, local server setup, useful integrations, tool design, permissions, and troubleshooting.

MCP · AI Tools · Tool Integration · Developer Tools

Engineering & Leadership

January 28, 2025 · 6 min read

What Contributing to Drupal Taught Me

A real patch review, the work around a contribution, and how to make yourself useful without taking on too much.

Open Source · Drupal · Community · Contributing

AI & Agents

January 27, 2025 · 6 min read

AI Agents in the Enterprise: Where to Start

A practical approach to enterprise AI agents: tool access, memory, orchestration, evaluation, and a bounded first deployment.

AI Agents · Enterprise AI · Tool Calling · Evaluation

Frontend & Web

January 25, 2025 · 8 min read

Frontend Development: Which Changes Are Worth Your Time?

How to evaluate frontend changes without chasing every trend: AI assistance, signals, WebAssembly, rendering, accessibility, and development tools.

Frontend · AI Development · Angular · WebAssembly

Frontend & Web

January 20, 2025 · 7 min read

Web Performance: Find the Wait, Then Fix It

A practical method for finding slow user journeys, tracing the cause, and checking that a performance change actually helps.

Performance · Web Vitals · Frontend · Architecture

Engineering & Leadership

January 15, 2025 · 6 min read

Drupal: Where It Fits and What It Costs

Content models, editorial workflows, and the tradeoffs of a separate frontend, from an architect who worked with Drupal at Acquia.

Drupal · CMS · Architecture · Acquia

Frontend & Web

December 15, 2024 · 6 min read

Building Angular Apps That Stay Easy to Change

Feature boundaries, state ownership, and rendering choices that keep a growing Angular application understandable.

Angular · TypeScript · Architecture · Performance

Engineering & Leadership

December 1, 2024 · 6 min read

Leading Engineering Teams as They Grow

Clear ownership, useful processes, and room for people to make decisions as a team grows.

Leadership · Engineering Management · Team Building

Frontend & Web

November 28, 2024 · 7 min read

JavaScript from ES6 to ES2024: What Actually Helps

A practical look at modern JavaScript: clearer async code, safer defaults, and data transformations with fewer surprises.

JavaScript · ES2024 · Web Development · Frontend

Conversation

Let's Discuss Technology

Disagree with something I wrote, or wrestling with a similar problem? I read every reply and usually learn something from them.