Decisions
The standing decisions every other page interprets — plus the items still open.
This page records the standing decisions that govern the program. Every other page interprets these. When a page and this list disagree, this list wins.
Standing decisions
| Decision | What it means in practice |
|---|---|
| We charge for AI-accelerated products and services | No free pilots of paid capability; proof engagements are priced engagements |
| Services over products — "close the books, don't sell QuickBooks" | We sell outcomes delivered through the platform, not shrink-wrapped software |
| X10 is the subscription and hosting surface | Every product ships through X10 auth, billing, and audit |
| Trust Zones decide what gets automated | Automate the high-volume, low-stakes, reversible work; everything else keeps a human in the loop |
| X10 Dividend is the commercial spearhead | The savings assessment opens every account; the three-stage product converts it to recurring revenue |
| The assessment is priced at 25% of realized annual savings | The customer pays a share of measured recovery — see the benchmarks |
| Vizzy subscriptions are credit-based | Plans start at 20 credits per month; one credit covers infrastructure and model cost |
| The platform CLIs are the delivery engine | vct-cribl-cli and vct-splunk-cli own the integration layer for both platforms |
| The knowledge base is the durable competitive advantage | Protected while cooperating with Cribl, Splunk, and CrowdStrike; tools are commodities, the corpus is not |
| Demo polish outranks back-end completeness | A working demonstration sells; an invisible back end does not |
| AI development directly on the X10 codebase is paused | In force until the security assessment completes — X10 holds production credentials |
| AI output is reviewed before it ships | Every generated line of code and every proposed change is engineer-approved; vendor claims are independently validated |
| Clients own delivered solutions; VisiCore retains the knowledge | The IP frame behind the knowledge base |
| Training data is generated by VisiCore's own systems | Completely independent of customer data |
The real gate
Customer compliance and consent stand between the products and revenue. Technology does not. The tooling is built and demonstrated. But enterprise customers first need answers on data boundaries, consent, and code security before they deploy. Enterprise AI governance is strict. Business units approve on their own. Timelines run months to years per vendor. Reviews can top a hundred questions. The plan assumes a full review cycle. The no-AI core of X10 Dividend exists to clear it: a scan with no AI in the path needs no AI approval. The same input always gives the same output.
Open items
These items need a human decision or action.
| # | Open item | Owner | Notes |
|---|---|---|---|
| 1 | "Guaranteed X% reduction" language | Rob + James | Requires a contractual guarantee construct; until then all claims stay measured, not promised |
| 2 | Code-security sign-off for X10 endpoints | Paul | On the beta-gate path; also lifts the X10-codebase pause |
| 3 | Data Access Policy | Rob | Sequenced with the first client engagement |
| 4 | Benchmark gaps: Windows events, syslog, F5 LTM, Cisco Umbrella, VMware | Jacob + Vince | One before/after measurement each |
| 5 | Cribl GOAT nominations | Rob + Paul | Submission deadline 2026-07-29 |
| 6 | Confirm the deployed docs-site URL and access policy | Jacob | Cloudflare hosting per docs/cloudflare_hosting_guide.md |
Ideas without a mandate
Token Tracker, GoatOptimize, Medallion Mapper, Config Audit Bot, Prompt Firewall, GoatRoute, Runbook Generator, CIS Compliance Whisperer, Collector Forge, SOW Accelerator, and Telemetry Data Co-op are scored backlog candidates, each awaiting a real decision. See the idea pipeline.