Bitlemmas University Koans Let’s talk

Build with AI.
Test and explain
every step.

Bitlemmas Agentic Systems Week

Five days (40 hours) of personal, one-on-one training with our instructor for executives, contractors, and project and product owners.

Discuss your 1-on-1 training week

Work one-on-one with our instructor, presenter of the hands-on 2024 SXSW open-source 5G workshop listed in Forbes’s SXSW guide ↗.

Executives & process owners Contractors & consultants Project & product owners

Professional training · Request the detailed curriculum →

An AI agent is a program that takes steps and uses tools under rules you set. You learn through University Koans, our curriculum of short, hands-on exercises based on requirements for a university admissions system—handling applications, required documents, staff reviews, and admission decisions. In the later part of the week, your instructor tailors the examples to one focused part of the agent or workflow you choose to make, so the rules and judgment you use in your work become the rules your agent follows.

401-on-1 hours

5focused days

3role paths

1-on-1 coaching.
Tailored to your role.

How the training teaches

Start with clear requirements.
Build an agent step by step.

You learn the method on our admissions practice example first, then apply the same building blocks to a focused part of the agent or workflow you want to make.

Map the requirements in plain English

In the first lesson, read the written admissions requirements, mark the actions and rules, and map out the workflow stages before any code is written.

Run repeated AI prompt tests

Test a short prompt across repeated live runs to see where an AI model guesses, then measure what improves as you add clear rules.

Build rules, tools & an agentic observer

Pair program—working together on the same code with your instructor—to build stage rules, allowed actions, diagrams, and an agentic observer (an observer part you build into your agent to see what step it is on, what action was tried, and what rules allowed or blocked it).

Shape the agent you choose to make

In the later sessions, your instructor tailors the examples to a focused part of your own agent so the rules, exceptions, and decisions from your work guide how your agent operates.

See how the training & week work

Your machine & models

Train on your own machine.
Windows, Mac, or Linux.

Bring the computer and AI model accounts you already use. We prepare and check your setup before Day 1.

In a 60–90 minute setup session before Day 1, we get your Windows, Mac, or Linux computer ready with a Linux coding workspace (on your computer or on a private remote Linux computer). We also check the AI model and coding assistant connections you want to use for prompt experiments, provided workshop chat, and hands-on coding.

See machine setup & supported models

Choose your role path

Three ways through the week.
Coached 1-on-1 for your role.

Each week is hands-on and one-on-one, so we match the pace to your starting point. Before Day 1, we prepare your setup and agree on one focused part of your agent or workflow.

Build one part of a business agent with AI tools

Turn your excitement about Claude Code and AI coding tools into a tested part of a business agent.

You already use AI coding tools—now use them with one-on-one coaching to build one focused part of an agent for your business. Pair program with our instructor so the rules, exceptions, and decisions you know from your work guide the code, and test every change instead of running on autopilot.

  • Turn your business rules, allowed actions, and exception handling into working code and automated checks.
  • Use Claude Code or your preferred supported coding assistant with direct coaching to add stage rules, diagrams, and an observer part.
  • Inspect, test, and explain every change so you leave with a working prototype part and a clear next-steps plan for your business.

No deep programming background required. We set up your workspace before Day 1 and coach you hands-on as you code with AI tools using sample business records.

Talk with us about your role

Your instructor

1-on-1 with W. Watson.
Personal coaching & pair programming.

You train one-on-one directly with W. Watson, who walks you through each exercise and pair programs with you on the same code, just as he has done with dozens of people throughout his career. He builds open-source and cloud software with Vulk and Bitlemmas, focusing on automated checks and repeatable test setups that verify whether real systems follow stated rules.

5G & mobile networks

Telecommunications

As mentioned in Forbes’s SXSW 2024 guide ↗—named in the article introduction and listed as “Build Your Own Open Source 5G Tower Workshop.”

Presented the hands-on 2024 SXSW open-source 5G workshop with Vulk (workshop guide ↗, presenter slides ↗), building on a live 5G mobile network call demonstrated on stage at Cloud Native Rejekts NA 2023 ↗.

Cloud-native & open source

Cloud

Agentic systems & koans

AI

  • Open-source University Koans curriculum: Created our hands-on exercises and practice workspace for building and testing AI agents with stage rules, allowed actions, visual diagrams, and an agentic observer.
  • 1-on-1 AI agent training & pair programming: This training week pairs you one-on-one with your instructor to build and check one focused part of an agent using supported AI models and coding assistants →.

Inquire

Tell us where
you’re starting.

Prepare an email draft here to ask about a 1-on-1 training week or request the detailed curriculum. Once you send your email, we reply within two business days (and send the detailed curriculum if you requested it).

info@bitlemmas.com

Training is one-on-one with W. Watson (ask us if you want to arrange a custom team week). You can also connect in the Bitlemmas Discord ↗.