Bayesian Leaf

Link copied.

Brick & Mortar

Your tools don't talk to each other. Your team does it for them.

Your operations and logistics run on too many apps. The day goes to switching between them, and every switch costs margin.

Atlanta local. Helpful anywhere.

Start with the workflow that costs you the most. In a week, you'll see where the time goes, and one change working.

The work is tailored to your business: the tools you already run on, and the way your team works today.

Where it starts

Pick one to see where it happens.

Who it's for

The business

A growing business, often opening another location or taking on more work than its systems were built for.

Who I talk to

The owner, CEO, COO, or head of operations: whoever answers for how the business runs day to day.

What I actually do

  1. Map the operation The tools, the data, and how a normal week runs, drawn in one place.
  2. Find the load-bearing decision One decision holds up the rest.
  3. Make it legible Whoever acts on a tool's call should see what it decided and why.
  4. Build the trust layer The reason behind each call, an override, and a way back when the tool is wrong.
  5. Ground it in measurement Pick the number that should move, hours or margin, before anything changes.
  6. Sequence it Pilot it in one place first, then roll it out to the rest.

Where I've done this

At Opendoor, I co-led research with the operations teams, thirty-five sessions with the people doing the work, and co-authored a twelve-capability blueprint for the internal tools they work in.

I also led design on Pinterest's automated ad product and designed audience tools for Meta's advertisers. Every engagement is with me, start to finish.

The first engagement

Operations ReadOne week

$9K

  • A map of one bounded workflow: the tools, the data, and the people in it.
  • The consequential decision in that workflow, and who owns it.
  • A written account of where the tools and the people disagree.
  • A focused prototype or demonstration of one proposed change.

We pick the workflow, and the access it needs, in the working session before the read starts. The prototype shows the change; building it into your live tools comes after the read, as a Direction Sprint or a Standing Engagement.

What comes after the read
Direction SprintFour weeks, fixed scope. $25K.

A direction document with named principles your team can push back on, a working prototype with the direction inside it, and the sequence to roll it out.

Standing EngagementMonthly. $12K to $20K/mo.

I work alongside the person accountable for operations, and can carry a defined product or workflow change through design and implementation. $12K holds the direction. Adding hands-on work moves it toward $20K, and back down when nothing needs building.

AdvisoryAfter an engagement. $1,600/mo.

A bookable weekly session of up to an hour. It keeps the context current.

How it starts

A 30-minute working session. No deck, no pitch. Bring whatever is stuck. We work out what's going on, whether it's work I do, and where to start.

The practice

Bayesian Leaf is my practice. Most of it serves companies building AI products. This page is for the businesses those products are built for.

See the practice, or read its published case: an applied AI platform, not an operations engagement.

Contact

hew@bayesianleaf.co bayesianleaf.co Atlanta, GA, working worldwide
Book a working session