Revenue Architecture.

Kindo captures a portion of Deloitte's current Swimlane expenditure. The value proposition: Deloitte replaces a legacy SOAR platform with a platform that also provides agentic capabilities (triage agent already in production), at equal or lower total cost, while eliminating the integration seam between Swimlane and Kindo.

GAP — NOT DOCUMENTED
Swimlane contract value (annual license + support) is not documented in available materials. Deloitte's internal cost of running Swimlane — including Harshal and Archith's continuous playbook maintenance load (recorded in the July 27 Discovery session as a constant engineering commitment) — is also undocumented. Both figures are required to size this mechanism. Without them, the commercial anchor has no number attached.

Increased Kindo platform fee or utilization fee that absorbs a negotiated share of the displaced Swimlane spend. Deloitte should pay less than Swimlane cost (they retain savings) and more than current Kindo fees (Kindo captures value). The exact split is a negotiation — the available figure determines the range.

Tony (Kindo-side positioning) + Deloitte procurement (contract terms). Kush/Krishna own the internal business case for Deloitte.

• Turbo Mode (or equivalent deterministic layer) must be production-ready before the December go/no-go — D1 unresolved
• Enablement tooling must exist for Deloitte to migrate independently — migration tooling and visual workflow builder are on the critical path
• Kindo must be able to handle the alert volume currently flowing through Swimlane — volume figures undocumented
• The Swimlane contract value must be retrieved to set a negotiation range

The deterministic layer reduces token spend per alert — this is precisely Zun's stated concern (July 27): "we do not want everything to be token-based; some very easy things, if API can do it, why do we need to burn tokens?" Unit price falls. But volume rises substantially because Kindo now processes 100% of alerts, not just the subset that Swimlane forwards to the triage agent. The net effect — lower unit cost × higher volume — may be the most commercially valuable element of the entire replacement. Nothing in the material to date has treated it as such.

GAP — VOLUME FIGURES NOT DOCUMENTED
Alert volume, concurrent execution count, and peak load were not provided in the July 27 Discovery session (Zun said the load "can be very very high"). These figures are required to model the net consumption effect. Without them: if the deterministic layer reduces per-alert token cost by X% and total alert volume is Y× what Kindo currently processes, the net consumption change is calculable — but X and Y are both unknown.

GAP — CURRENT KINDO CONSUMPTION BASELINE NOT DOCUMENTED
What Deloitte currently pays Kindo in consumption/utilization fees for the triage agent is also required to calculate the delta.

Consumption-based pricing (tokens, agent runs, or alert volume). The deterministic layer should be priced differently from the agentic layer — deterministic processing is cheaper to run and should be priced to encourage migration from Swimlane, not penalize volume. A blended rate or tiered model may apply.

Tony (pricing structure) + Charlie (capacity/cost modeling) + Kush/Krishna (volume commitment).

• Turbo Mode must handle Deloitte's alert volume at scale — 80% of their clients currently route through Swimlane
• The per-alert cost in deterministic mode must be materially lower than current agentic processing — otherwise volume expansion is a cost center, not revenue
• Alert volume, peak load, and current consumption baseline must be retrieved from Deloitte to model the net effect
• Pricing model for deterministic vs. agentic processing must be defined

Revenue from jointly developing SOC for AI capabilities that Kindo can then sell to the broader market. Deloitte contributes domain expertise, SOC operational requirements, and validation. Kindo contributes platform engineering and AI capability. Commercial model is open — see three options below.

Available reference points:
Alliance net new revenue (A.6–A.13): $1–2M+ estimated. Source: program planning materials.
Per-agent pricing range: $2,000–$8,000/month depending on agent type. Source: May 19 Tony-Joana call.
Full deployment projection (6 agents × 100 installs × $3,500/month blended): $25.2M/year. Source: May 19 revenue model.
SOC for AI has replaced deprioritized agents as the lead net new revenue play.

GAP — SOC FOR AI SPECIFIC REVENUE NOT MODELED
These figures are portfolio-level estimates, not SOC for AI-specific. No dedicated revenue model for SOC for AI co-development exists. The magnitude depends entirely on D6 (commercial model) — development fee, co-development, or revenue participation each produce different numbers.

Three models. No position adopted. No recommendation.

D6 — unresolved. Tony and Ron on Kindo side. Kush on Deloitte side.

• D6 must be decided before the next co-design session — four Sprint 2 items are actively absorbing Deloitte design input now
• Co-development IP terms must be established — currently no agreement exists (evidence gap)
• What was agreed with Krishna and Kush about ownership of jointly designed capabilities must be documented — currently undocumented
• Track 2 deliverables must be enumerated and classified by contribution (Kindo-initiated vs. Deloitte-shaped) — the D5 split provides a starting framework