In May 2026, VTKL walked into Deloitte's Cyber Operate headquarters with a bold claim: we can take your $300M/year security practice from 40% EBITDA to 80%. Not by adding headcount. Not by cutting corners. By building AI agents that learn — and compound — every single day they're in production.
Deloitte's team was already running at 70–85% efficiency before any AI. The drag was underneath: Swimlane licensing at $3–6M/year, CrowdStrike markups at $2–4M, and headcount that couldn't scale without adding bodies. Tony saw what nobody else was framing: the unlock was structural, not operational.
Deloitte's $5.5M annual contract pays for every agent built. R&D cost to Kindo: zero. Once those agents are battle-tested in Fortune 100 production, Kindo sells them to the market at 2–5× revenue multiplier. The alliance and the market aren't separate businesses — they're two ends of the same flywheel.
"Every agent is a net new revenue goal. Either it brings new revenue dollars, or it drives better profit margins."
— Krishna, D&RaaS Lead · May 7, 2026Kush defined institutional knowledge as compound learning through use — not a document, not a one-time transfer. Every alert resolved teaches the agent. Week ten is smarter than week six. The competitor who starts six months later doesn't just start later — they start behind on a curve that's accelerating.
The four agents in production today represent roughly 5% of the total agent opportunity across Deloitte's Cyber Operate portfolio. Five of six service lines have zero agent coverage. D&RaaS is the beachhead. Every other service line follows the same playbook: one relationship, one IK capture, one agent package — then it scales.
Tony built a three-tier packaging model from Kush's own words. Each tier earns the right to sell the next. They are not separate products — they are a compounding stack where every deployment at the lower tier creates the proof and the patterns for the tier above.
For every alliance dollar earned from net new agents at Deloitte, the Mythos market produces 2–5× in revenue. The R&D is already paid for by Deloitte. Every dollar of market revenue is near-pure margin. Total combined opportunity: $33M at the floor, $99M at the ceiling.
Seven was the right number to say in the room. But the work requires nine — and the two extra aren't generic engineers. They are security analysts and consultants who design agent workflows from the inside. People who understand D&RaaS or Identity deeply enough to configure the logic, not just build the software. 60% of this team will be Agent Designers.
"Tony is the only person at Kindo who can acquire the institutional knowledge that 55% of the scope and 60–75% of the EBITDA target depend on."
— Tony's CDO Thesis · May 22, 2026HP is the first Fortune 100 production client — four phases from installation through fully autonomous steady-state operation. As of today: 1 of 100 production equivalents complete. Deloitte deploys to clients before internal — no internal proving ground. Every deployment carries full production weight from day one.
The plan is strong. But three items could derail the timeline if not tracked actively this week.
Steve Ward — tech consultant, PE observer, Trent Johnson's close friend — tested Warren on May 25. He came expecting a chatbot. He left calling it "Jarvis, not Siri." A decision-making operating system. PE observers were floored. And the key insight surfaced: Tony's 300–400 hours of executive coaching transcripts are the secret weapon — the corpus for a full Digital CEO twin.
"Warren's active listening gives pointed answers — versus ChatGPT and Claude giving lots of different solutions."
— Steve Ward · May 25, 2026Deloitte funds the build. Mythos creates the market. Tony holds the institutional knowledge no one else can acquire at the speed this opportunity demands. Warren executes at the pace the window requires. The flywheel is live. The window is open. The question now is how fast the team moves.
Prepared by Aria · VTKL · May 27, 2026