What is an AI-native company?
An AI-native company redesigns workflows, decisions, data access and software so AI is part of how the business runs. What it means, how it differs from adding AI tools, and how to start.
Read the articleInsights
Insights on AI Transformation, AI Agents and Product Building
Research and operating notes. From the work itself
In-depth pieces on AI-native operations, agents, product builds and the ventures we create. Original frameworks and the reasoning behind them, written by the people doing the work.
Latest insights
Don't add AI to a broken process: map the workflow first
Automating a workflow you haven't mapped usually automates its problems. A practical, step-by-step way to map a business process before writing any code or adding AI.
Read the articleWhat should an AI agent be allowed to do? A practical model
A practical permission model for AI agents: four levels of action from read-only to high-impact, how to enforce them in code rather than prompts, and how to design the handoff to people.
Read the articleForward deployed engineering, explained
What a forward deployed engineer does, why the role matters for AI implementation, and how it differs from consulting, outsourced development and on-site support.
Read the articleThe anatomy of a production AI agent
Knowledge, tools, permissions, a provider abstraction, evaluation and observability: the parts that turn an AI demo into a production AI agent a business can rely on.
Read the articleAgentic AI in 2026: what actually changes for operations teams
Agents that read, decide and act inside business systems are moving from pilots into operations. What genuinely changed in 2026, what has not, and how operations teams should respond.
Read the articleVoice AI agents for business: design rules that hold up
Opt-in, live transcripts, interruption, short turns and a text fallback. Practical design rules for voice AI agents that customers and staff will actually trust.
Read the article
Operating Notes
How real workflows change.
What is an AI-native company?
An AI-native company redesigns workflows, decisions, data access and software so AI is part of how the business runs. What it means, how it differs from adding AI tools, and how to start.
Read the articleDon't add AI to a broken process: map the workflow first
Automating a workflow you haven't mapped usually automates its problems. A practical, step-by-step way to map a business process before writing any code or adding AI.
Read the article
Agent Patterns
Practical agent architectures and controls.
What should an AI agent be allowed to do? A practical model
A practical permission model for AI agents: four levels of action from read-only to high-impact, how to enforce them in code rather than prompts, and how to design the handoff to people.
Read the articleVoice AI agents for business: design rules that hold up
Opt-in, live transcripts, interruption, short turns and a text fallback. Practical design rules for voice AI agents that customers and staff will actually trust.
Read the articleBuild or buy? Deciding on a custom AI agent
When an off-the-shelf AI tool is enough, when a custom AI agent is worth building, the questions that decide it, and the middle path most good systems end up taking.
Read the article
Build Notes
Technical lessons from products.
The anatomy of a production AI agent
Knowledge, tools, permissions, a provider abstraction, evaluation and observability: the parts that turn an AI demo into a production AI agent a business can rely on.
Read the articleWhat drives the cost of an AI agent project?
The factors that actually determine the build effort and running cost of an AI agent: integrations, data readiness, actions and permissions, evaluation, interfaces and volume, and how to keep it economical.
Read the article
Field Notes
Observations from discovery and deployment.
Forward deployed engineering, explained
What a forward deployed engineer does, why the role matters for AI implementation, and how it differs from consulting, outsourced development and on-site support.
Read the articleAgentic AI in 2026: what actually changes for operations teams
Agents that read, decide and act inside business systems are moving from pilots into operations. What genuinely changed in 2026, what has not, and how operations teams should respond.
Read the articleAI agency or AI transformation partner? How to choose
What separates an AI agency that ships features from an AI transformation partner that changes how your company operates, the questions that reveal the difference, and how to choose.
Read the article
Venture Thesis
Problems worth creating companies around.
LoopGrid: why cities need proof, not just pickups
The thesis behind LoopGrid, municipal waste-management software that connects collection tracking, digital manifests and verified destinations to measurable recovery and landfill diversion.
Read the articleAuraCampus: teaching intelligence that stays accountable to teachers
Why AuraCampus turns approved curriculum into visual lessons and teacher-reviewed learning evidence, with teachers in control of the AI and meaningful updates for parents.
Read the articleThroughline: why a resume should start with your story
The thesis behind Throughline, career software that builds role-specific resumes from a long-term story vault and job-posting analysis instead of rewriting an old document.
Read the articleHeroMode: designing for the comeback, not the streak
Why HeroMode, a discipline RPG, is built around the moment after a missed day: short daily quests, proof-based progress and comeback missions instead of fragile streaks.
Read the articleMirra: building a private place to reflect
How Mirra approaches personal reflection software with structure, pattern awareness and visible privacy boundaries, and why it is designed to support reflection rather than diagnosis.
Read the articleMiyaar: one line of accountability from design to handover
Why Miyaar, a Bengaluru design-build practice for homes, interiors and brand environments, keeps architecture, engineering, procurement and construction within one accountable team.
Read the article
Get new notes by email
One email a month at most. Practical, from real work. Unsubscribe in one click.
Prefer a feed reader? Subscribe by RSS.