Learn to use AI in your program, from a coach who builds with it every day.

A working session with your staff, on your own data, on the jobs that actually eat your week. Not a talk about where the technology is going. You leave with things you can run on Monday and a written playbook for the ones you cannot yet.

Baseball AI consulting for college coaches.

Head coaches, assistants and operations staff at NCAA Division II, Division III, NAIA and junior college (JUCO) programs, baseball and softball. Nobody needs a technical background. If your staff can use a spreadsheet and a shared drive, that is the whole prerequisite.

It is aimed at the same kind of program the reports are: no analytics department, no operations budget, and a coaching staff already doing four jobs each.

What it covers

Concrete, in your own files, in this order or whatever order your staff needs:

• Turning a stat export into a scouting page. Take the CSV your program already downloads and get to a page a staff can read, without copying numbers by hand at eleven at night.
• Recruiting and parent communication. Drafting faster and better: first contact, follow up, the note after a visit, the message to a parent that has to be right the first time. Your voice, kept, not replaced.
• Practice plans and individual development plans from a template. One good template, then a plan per player that a coach edits instead of writes.
• Summarizing video notes. Turning what a coach dictated in the truck on the way home into something the rest of the staff can read.
• Building a simple opponent database. Where to keep what you learn about a team so it is still there next season, and how to get it back out.
• What to trust and what to check. The most important hour of the session. Numbers come from code, prose comes from the model, and a person signs off before it goes to a player or a parent. Where these tools are confidently wrong, and how to catch it.

Bench Coach

Bench Coach is the staff assistant this work points at: fed a program's own roster, schedule and scouting folder, it answers the questions a staff keeps having to look up. Every number it gives back is computed from those files, and a coach approves anything that leaves the building.

See Bench Coach.

The working model is public

The rules this work runs on are in a public repository: scouting-graph on GitHub. It is a working model on synthetic data. It is not how a staff gets a weekend packet, and it is not something to clone and load a real roster into.

College staffs who want reports, Bench Coach, or help standing any of this up start at Get started. Email from that page. Name the school, the sport, and the level.

How it works

• A remote session with your staff, 60 or 90 minutes, or a half-day workshop if you want to get through all of it and build as you go.
• Hands-on with your own data. Send exports and a couple of real jobs ahead of time and we work on those, not on a demo dataset.
• A written playbook afterward. The steps, the prompts and the checks, in a document your staff can hand to the next assistant.
• An optional monthly check-in for a staff that wants to keep going.

Who is teaching it

Matt Reichenbach. Computer engineer by training, more than twenty years building data and automation systems that run in production for businesses, and building production AI workflows now. Youth baseball coach, IHSA certified. Runs YourDataScouting, which delivers scouting and pitching reports to college staffs in season, so the examples in the session are the same work, not a case study from another industry.

What it costs

There is no fee for this right now. It is getting started the way the reports did: on real staffs, at no cost, while it proves useful.