Undeniable Reliabilly
KYBERNET · 2026-05-02
📋 Recap
  • Kybernet is an early-stage IT/networking startup founded by Drew Webb and William Klein, currently operating without a formal LLC or bank account.
  • The primary active client is Carl, with a completed Red Barn network installation and a pending categorized invoice of approximately $4,759 covering cost overruns.
  • The business has pivoted multiple times in recent weeks, most recently committing to a containerized private AI server product targeting HIPAA/SOC2-compliant SMBs in the commercial HVAC vertical.
  • Revenue remains minimal and inconsistent, with founders still relying on personal income and planning a pay-plus-reinvestment model to fund operations.
  • Business formation fundamentals — LLC registration, EIN, business bank account, and payment processing — remain incomplete despite being action items since late 2025.
📅 Agenda
  1. AI gateway product demo review and go-to-market readiness
  2. HVAC sector outreach strategy and first target prospects
  3. LLM gateway research recap (Nemo and Lera findings)
  4. LLC formation and EIN status update
  5. Carl project billing resolution and outstanding invoice follow-up
  6. Hardware standards and scope boundaries for AI server offering
  7. Python/Supabase agent build progress check-in
✅ Action Items
[William Klein] Build several software agents, focusing on solving a personal existing problem — no deadline mentioned
[Drew Webb] Configure agents to communicate with and request information from each other (agent orchestration) — no deadline mentioned
The group: Research LLM gateways (Nemo and Lera) and understand how large businesses implement secure AI solutions to scale down for SMBs — due before Thursday's meeting.
William Klein: Research HIPAA and SOC2 compliance, focusing on AI ethics and technology transference; prepare brief talking points for discussion on Thursday.
Drew Webb: Research the specific LLM gateways Nemo and Lera — due before Thursday's meeting.
❓ Questions
How should the proprietary HVAC knowledge base be built — through an interview process with the business owner, a top-10 frustrating call issues list, or automated monitoring of technician interactions?
How can a sequential Python script contain an agent that dynamically determines which other agent to engage for a given query?
How will permission access and terminal configuration challenges be managed when deploying agents that need to audit company data?
What is the exact pricing model for the HVAC agent product and the broader Kybernet service offering?
How will Kybernet package and redeploy agent builds across different clients using Docker?
📝 Notes
⭐ Highlights