Blog
Notes from the office.
Essays on how delegation should work when the account is real. No announcements dressed up as insight, and nothing we have not thought about properly.
Latest essay
AI agent management: Run agents like a team
Learn to manage AI agents with clear briefs, spending limits, weekly log reviews, and a practical way to retire agents that no longer earn their place.
Product · 22 October 2026 · 7 min read
Read the essayThe archive
Everything we have written.
Product · 21 Oct 2026 · 7 minAI agent templates: why they beat blank promptsLearn what good AI agent templates include, how they set scope, tools, budgets and approval rules, and when to build a template of your own.Read the essay 
Design · 21 Oct 2026 · 7 minMCP and AI agents: A guide for business operatorsLearn how MCP connects AI agents to business tools, why access scopes matter, and how to set approval rules before agents change records or contact people.Read the essay 
Design · 20 Oct 2026 · 6 minWhat to look for in an agentic AI platformLearn how to test an agentic AI platform for task execution, controlled permissions, human decision boundaries and reliable records before trusting a demo.Read the essay 
Product · 19 Oct 2026 · 7 minAI meeting assistant: what to check before a callLearn how to set up an AI meeting assistant that prepares a useful pre-call brief, checks past decisions and flags gaps before you join.Read the essay 
Product · 19 Oct 2026 · 6 minContent repurposing with an AI agentLearn how to brief an AI agent to turn one source into channel-specific drafts, preserve your brand voice, and review every piece before publication.Read the essay 
Trust · 18 Oct 2026 · 8 minAI resume screening with evidence behind every scoreLearn to build an AI resume screening process around written criteria, test for bias, inspect evidence and keep every rejection in human hands.Read the essay 
Product · 17 Oct 2026 · 8 minAI sales agent: Draft the pitch, let a person sendLearn how an AI sales agent can research prospects, draft personal outreach and track follow-ups while a person approves every first contact.Read the essay 
Trust · 16 Oct 2026 · 7 minAI bookkeeping: What an agent should never do aloneLearn which AI bookkeeping tasks are safe to delegate, why payments and refunds need approval, and how to keep a person in control of month-end close.Read the essay 
Product · 16 Oct 2026 · 5 minAI competitor analysis without invented numbersLearn how to use an AI agent for competitor research while keeping prices, plan details, product claims and changes tied to evidence.Read the essay 
Product · 15 Oct 2026 · 7 minAI market research: How to brief an agentLearn how to brief an AI research agent, choose credible sources, structure a useful market memo and catch invented claims before they shape decisions.Read the essay 
Founders · 14 Oct 2026 · 4 minAn AI executive assistant that asks before it sendsLearn what to delegate to an AI executive assistant, from inbox sorting to meeting briefs, while keeping approval over every important action.Read the essay 
Product · 14 Oct 2026 · 4 minInbox triage with an AI agent, step by stepLearn a practical inbox triage process with an AI agent, including labels, draft replies, escalations and a weekly review.Read the essay 
Product · 13 Oct 2026 · 8 minAI for small business: The first three jobs to hand overA practical order for giving recurring work to AI agents, checking what comes back, and keeping customer promises and financial decisions in human hands.Read the essay 
Founders · 13 Oct 2026 · 7 minAI tools for startups: When to use an agentLearn which startup tasks need an AI assistant, which deserve an agent, and how to choose without adding supervision work to a small team.Read the essay 
Design · 12 Oct 2026 · 7 minWhat makes AI teammates different from toolsLearn how AI teammates take assignments, return finished work and change the way teams handle handoffs, permissions and human review.Read the essay 
Product · 11 Oct 2026 · 7 minHow to plan an AI workforce for a small teamLearn how to build an AI workforce around two recurring tasks, set budgets and approval rules, and decide what to keep after a 30-day trial.Read the essay 
Design · 10 Oct 2026 · 7 minAI employees: Where the metaphor breaksAI employees can take on assigned work, but they cannot hold responsibility. Learn where the employee comparison helps and where it creates risk.Read the essay 
Trust · 9 Oct 2026 · 7 minAI agent security: 12 questions to ask a vendorA practical AI agent security checklist covering permissions, secrets, human approvals, audit logs, data retention and emergency shutdown.Read the essay 
Trust · 8 Oct 2026 · 7 minWhat a useful AI audit trail must recordLearn what an AI audit trail should record, how to verify approvals and payloads, and why an editable activity log cannot provide reliable evidence.Read the essay 
Agencies · 7 Oct 2026 · 6 minWhite label AI agents: what agencies should checkA practical checklist for agencies assessing white label AI agents, covering liability, client visibility, data, approvals and operating control.Read the essay 
Agencies · 6 Oct 2026 · 6 minAI for marketing agencies without losing client controlHow agencies can run AI agents on client work with separate workspaces, approval gates, clear budgets and a record clients can inspect.Read the essay 
Agencies · 5 Oct 2026 · 5 minAI for consultants: hand over prep, keep judgementA practical split of consulting work: what AI agents can prepare, what still needs a consultant, and where human review should sit.Read the essay 
Design · 4 Oct 2026 · 5 minAn AI task manager is not an AI agentLearn what an AI task manager tracks, what an AI agent actually does, and how to choose the right tool for the work you want off your plate.Read the essay 
Product · 3 Oct 2026 · 6 minWhat AI agents can do today, and where they stopA practical map of work AI agents handle well, the tasks that still need human judgement, and how to decide what to delegate.Read the essay 
Product · 2 Oct 2026 · 6 minAI agent use cases that earn their keepA practical test for AI agent use cases: repetitive work, checkable output and a human checkpoint before mistakes leave the workspace.Read the essay 
Product · 1 Oct 2026 · 5 minNo code AI agents: the decisions you still makeBuilding no code AI agents removes programming work, but teams still decide scope, permissions, budgets, tools and approvals.Read the essay 
Trust · 30 Sept 2026 · 5 minAI agent governance for a team of tenA practical governance model for small teams: who creates agents, who approves risky actions, where logs live, and who reviews them.Read the essay 
Founders · 29 Sept 2026 · 4 minWhat an AI chief of staff can and cannot doA practical guide to what an AI chief of staff can handle today, where human judgement stays in charge, and how founders can delegate safely.Read the essay 
Trust · 29 Sept 2026 · 4 minAI guardrails that hold when agents are busyLearn why AI guardrails need enforced rules, approvals and logs instead of relying on an agent to remember instructions.Read the essay 
Trust · 29 Sept 2026 · 4 minHuman in the loop automation without the bottleneckLearn where people fit in human in the loop automation so AI agents move faster while important actions still receive a clear sign-off.Read the essay 
Product · 28 Sept 2026 · 7 minHow to delegate tasks to an AI agentLearn how to brief an AI agent, set a spending limit, define approval checkpoints and review actions before they affect people or records.Read the essay 
Trust · 19 Sept 2026 · 7 minWhat human in the loop means for AI agentsHuman in the loop is not a setting. Where the loop belongs for an AI agent, what makes a checkpoint real, and why silence must never count as a yes.Read the essay 
Product · 19 Sept 2026 · 6 minAutonomy is not the hard partEvery agent demo shows a long unattended run. The hard part is blast radius: what an AI agent may do alone, what waits for a person, and who pays for it.Read the essay 
Design · 19 Sept 2026 · 6 minAn AI agent is not a chatbotA chatbot answers; an AI agent acts. What changes when software can send, pay and publish, and what that means for how you brief it and supervise it.Read the essay 
Pricing · 16 Sept 2026 · 5 minWhat a delegation costsCost per step, budget caps and credits: how the bill stays visible while the work runs, and why a hidden bill is the fastest way to lose trust.Read the essay 
Trust · 26 Aug 2026 · 5 minEight things an agent never does without a personThe floor under every policy in Delagents: eight actions that always stop for a sign-off, whatever the agent's rules say, and the reasoning behind each.Read the essay 
Design · 4 Aug 2026 · 4 minA brief beats a flowchartBriefs describe the outcome; builders demand every branch before the work exists. Where each one wins, and why Delagents starts from the brief.Read the essay 
Product · 14 Jul 2026 · 5 minThe approval is the productWhy the approval card, and not the agent builder, is the center of a delegation product. Blast radius, not capability, decides what people hand over.Read the essay
The reading list
One essay at a time, when there is one.
New writing on delegation, sign-off and what it costs to let software act. No product announcements dressed up as insight, and no weekly email for the sake of a weekly email.