Live demo fleet Public demo surface with real backends and synthetic, public-safe data.

Demos · public proof surface

Open the software and see how it works.

Each demo here is a working tool with a real backend. The point is simple: show how I build AI, RevOps, and internal tooling systems under live constraints, with synthetic or public-safe data where the boundary requires it.

Open RevOps Software Factory Open fleet operations
Fleet size
7 demos
Live backends
7 running
Core lanes
4 focus areas
Public repos
7 sources

Featured demos

These four give a good sense of the range: one multi-agent review system, one revenue-flow debugger, one ops-code gate, and one run-inspection tool.

Applied AI · RevOps

RevOps Software Factory

Featured

Multi-Agent Revenue-Data Review

Multi-agent review for synthetic revenue data.

A coordinator dispatches specialist agents (dedup, attribution, stage-logic, enrichment) over a synthetic HubSpot batch, fuses their findings, and returns one approval-biased verdict with a live token-cost ledger. Risk-scaled compute, model tiering, and a cache + budget guard make the live backend safe and cheap. (Synthetic data, not a live CRM.)

  • Anthropic
  • HubSpot
  • Multi-agent
Open demo One-pager Source

Revenue systems

Q2See

Featured

Quote-to-Cash Flow Inspector

Inspect a quote-to-cash flow record by record.

Import your own quote-to-cash export and the backend parses it server-side into one quote→contract→invoice→renewal flow, flagging the broken handoff before revenue leaks. (A point-in-time export, not a live CRM connection.)

  • Salesforce
  • Stripe
  • Workato
Open demo One-pager Source

AI ops-code review

Apexlint

Featured

Ops-Code Linter

Deterministic review for Apex, Flow, and n8n changes.

Deterministic lint of agent-generated Apex / Flow / n8n before it hits production: SOQL-in-loop, missing bulkification, no error handling.

  • Salesforce
  • Apex
Open demo One-pager Source

AI observability

Tracewell

Featured

Agent Run Explorer

Inspect a failed agent run step by step.

Reads a failed agent run down to the token: tool-call timeline, cost per node, the prompt regression that caused it.

  • Agent traces
  • Observability
Open demo One-pager Source

All demos

Browse the demos by lane

Open one, use sample data, and see how it behaves. Each runs on its own backend, with a one-pager and a public source repo attached.

  1. 01

    RevOps Software Factory

    Featured

    Multi-Agent Revenue-Data Review

    Review a synthetic revenue batch with a multi-agent backend and live cost tracking.

    Real multi-agent backend

    • Anthropic
    • HubSpot
    • Multi-agent
    Open One-pager Source
  2. 02

    Q2See

    Featured

    Quote-to-Cash Flow Inspector

    Import a quote-to-cash export and see where the handoff broke.

    Real import backend

    • Salesforce
    • Stripe
    • Workato
    Open One-pager Source
  3. 03

    Apexlint

    Featured

    Ops-Code Linter

    Lint Apex, Flow JSON, and n8n DSL before it reaches production.

    Real linter backend

    • Salesforce
    • Apex
    Open One-pager Source
  4. 04

    Tracewell

    Featured

    Agent Run Explorer

    See the tool-call timeline, prompt diff, and token cost behind a failed run.

    Real analysis backend

    • Agent traces
    • Observability
    Open One-pager Source
  5. 05

    Funnelguard

    Funnel Config Linter

    Catch broken UTM taxonomy, orphaned forms, and lifecycle contradictions at config time.

    Real linter backend

    • HubSpot
    • GA4
    Open One-pager Source
  6. 06

    Forager

    Internal Knowledge Harvester

    Upload a Slack export and extract resolved threads into a searchable, confidence-scored knowledge base.

    Real import backend

    • Slack
    • MCP
    Open One-pager Source
  7. 07

    Inboxward

    Deliverability Cockpit

    Inspect SPF, DKIM, and DMARC live, then layer synthetic fleet scenarios on top.

    Live DNS backend

    • SPF
    • DKIM
    • DMARC
    Open One-pager Source

Open source

Headline public repos

Production extracts, not demos — clone and read the code directly. Full catalog at dallascrilley.com/proof.

Shipwright

An agent that turns approved GitHub issues into tested, reviewable pull requests.

Runs in a disposable sandbox; a PR only exists if the repo’s own verification command exits zero and a human says publish.

  • TypeScript
  • Bun
Source

Vouch

Human review as an API: agents submit work, real reviewers return a consensus verdict.

173 tests, 72 files, plus offline verification harnesses.

  • TypeScript
Source

Holdfast

An append-only decision ledger with a human approval gate: AI can propose, only a person can publish.

Enforced in Postgres by triggers and a hash chain; 38 tests, 14 of which attack the ledger directly.

  • TypeScript
  • Postgres
Source

Why this page exists

The work should be easy to inspect directly.

The fleet operations layer probes the live backends, shows the ship gates, and links each public repo. If the question is whether a demo is up and doing real work, that page is the fastest check.

Inspect the fleet