Notes from teams running on agents.

How real teams wire their AI together, written plainly. New posts most weeks.

Featured

Dana Whitfield

•

Sep 8, 2026

The handoff problem: why most agent pipelines fall apart in week one Copy

A short read on where multi-agent setups actually break, and the patterns that hold.

Orchestration

5 min read

Featured

Dana Whitfield

•

Sep 8, 2026

The handoff problem: why most agent pipelines fall apart in week one Copy

A short read on where multi-agent setups actually break, and the patterns that hold.

Orchestration

5 min read

Dana Whitfield

•

Sep 1, 2026

Checkpoints, not babysitting: where to let AI run and where to stay involved

The difference between reviewing every step and reviewing the moments that matter.

Orchestration

6 min read

Dana Whitfield

•

Sep 1, 2026

Checkpoints, not babysitting: where to let AI run and where to stay involved

The difference between reviewing every step and reviewing the moments that matter.

Orchestration

6 min read

Sam Ortega

•

Aug 25, 2026

We moved support escalations to an agent. We kept human.

What an escalation handoff actually looked like and why we kept humans.

Support

7 min read

Sam Ortega

•

Aug 4, 2026

The integrations you think you need vs the ones that matter

You'll be tempted to connect everything. Start with the four that actually move work.

Integrations

4 min read

Mara Chen

•

Aug 18, 2026

Picking your first workflow: start where the data already exists

The fastest way to a working agent isn't the flashiest workflow.

Getting started

4 min read

Mara Chen

•

Aug 11, 2026

Failures are data: reading your run log like a log, not a graveyard

Every failed run is a record of something your system couldn't handle. Here's how to read it.

Reliability

5 min read

Sam Ortega

•

Aug 25, 2026

We moved support escalations to an agent. We kept human.

What an escalation handoff actually looked like and why we kept humans.

Support

7 min read

Mara Chen

•

Aug 18, 2026

Picking your first workflow: start where the data already exists

The fastest way to a working agent isn't the flashiest workflow.

Getting started

4 min read

Mara Chen

•

Aug 11, 2026

Failures are data: reading your run log like a log, not a graveyard

Every failed run is a record of something your system couldn't handle. Here's how to read it.

Reliability

5 min read

Sam Ortega

•

Aug 4, 2026

The integrations you think you need vs the ones that matter

You'll be tempted to connect everything. Start with the four that actually move work.

Integrations

4 min read

Sam Ortega

•

Aug 25, 2026

We moved support escalations to an agent. We kept human.

What an escalation handoff actually looked like and why we kept humans.

Support

7 min read

Mara Chen

•

Aug 11, 2026

Failures are data: reading your run log like a log, not a graveyard

Every failed run is a record of something your system couldn't handle. Here's how to read it.

Reliability

5 min read

Mara Chen

•

Aug 18, 2026

Picking your first workflow: start where the data already exists

The fastest way to a working agent isn't the flashiest workflow.

Getting started

4 min read

Sam Ortega

•

Aug 4, 2026

The integrations you think you need vs the ones that matter

You'll be tempted to connect everything. Start with the four that actually move work.

Integrations

4 min read

Building your own agent stack?

We can show you.

Book a demo

Book a demo

Book a demo

Notes from teams running on agents.

The handoff problem: why most agent pipelines fall apart in week one Copy

The handoff problem: why most agent pipelines fall apart in week one Copy

Checkpoints, not babysitting: where to let AI run and where to stay involved

Checkpoints, not babysitting: where to let AI run and where to stay involved

We moved support escalations to an agent. We kept human.

The integrations you think you need vs the ones that matter

Picking your first workflow: start where the data already exists

Failures are data: reading your run log like a log, not a graveyard

We moved support escalations to an agent. We kept human.

Picking your first workflow: start where the data already exists

Failures are data: reading your run log like a log, not a graveyard

The integrations you think you need vs the ones that matter

We moved support escalations to an agent. We kept human.

Failures are data: reading your run log like a log, not a graveyard

Picking your first workflow: start where the data already exists

The integrations you think you need vs the ones that matter

Building your own agent stack?

Book a demo

Book a demo

Book a demo

Book a demo

Book a demo