Insights
Practical guides from our team.
What we’ve learned building AI and software for real businesses, and putting engineers on their teams.
Embedded teams
Hiring, vetting and working with engineers who join your team.
Staff augmentation vs. hiring vs. an agency: how to choose
You need more engineering capacity. You can hire, bring in embedded engineers, or hand the project to an agency. Each is the right answer in some situations and an expensive mistake in others. Here's how to tell which one you're in.
How to vet a remote software or AI engineer before they join your team
Remote hiring opened up a far bigger pool of engineers, and a far bigger pool of fake ones. Here's how to check that a candidate is real, can do the job and will fit your team, before they get access to anything.
What a bad engineering hire really costs, and how to lower the risk
The salary is the smallest part of it. A hire that doesn't work out costs months of recruiting, ramp-up, rework and a team that slowed down to carry them. Here's where the cost actually goes, and how to stop paying it.
Onboarding an embedded engineer: the first 30 days
A strong engineer with a weak first month looks like a weak engineer. This is the checklist we give team leads so an embedded engineer is shipping real work in their first week and fully part of the team by day 30.
AI in practice
Where AI pays off in a business, and how to do it safely.
Where to start with AI in a small or mid-sized business
Most AI projects stall because they start with the technology instead of the work. Start with one workflow your team does every day, measure it, and keep a person approving what matters.
Private AI assistants: keeping company data out of model training
An assistant that answers from your own contracts, email and records is only useful if you trust where that information goes. Here's how private assistants work, where the data lives, and what to ask before connecting anything.
How to tell if an AI project is worth building: a checklist
Seven questions to ask before you spend anything on an AI project, a simple way to score them, and what the score tells you to do next.
Checking flood insurance policies against the loan file with AI
Lenders have to confirm that every loan in a flood zone carries enough flood insurance, for as long as the loan runs. Here's what that check involves, why it's slow by hand, and how AI can do the reading while a person makes the call.
Building software
How good software gets from an idea to production.