Tools commoditize. Finished work with provenance does not.
Two former CTOs
We turn data into outcomes you can trace
CMD+RVL (Command Reveal) is an applied AI firm for finance and procurement teams. We turn questions that matter into finished, source-traced work your team can use, while building the information and context the next question will need.
There has never been more data within reach. Most teams already have more than they can account for, and another vendor ready to sell them more. AI does not solve that by itself: a model can give a traced number and an invented one with the same confidence. The hard part is knowing what holds up before the answer becomes a decision.
CMD+RVL closes that gap with finished work your team can inspect and use. The evidence stays attached, and what your team learns becomes the starting point for the next question.
Why fixed income first
Structured finance still runs on fragmented filings, disconnected figures, and workflows that are hard to reproduce. We started here because disclosure is abundant, the evidence is difficult to assemble, and a small sourcing error can change the answer. The tools finally exist. The rest is practice.
Bonds are a hard proof case, which is exactly why we picked them. We know the industry and we know the datasets. Our first proof case uses public SEC filings, so you can inspect the same source documents we used.
- Where every material number came from, down to the filing.
- What was true as of a given date, and what changed afterward.
- What answer would have been possible last quarter, reconstructible from evidence today.
The method travels anywhere a number has to survive the follow-up question, including procurement and supply chain.
How we work
We lead with the practice, not the platform. Most teams engage us around a defined outcome. When the capability belongs inside your firm, the CMD+RVL platform can also be licensed and deployed in your environment.
Show the gap.
We measure what can actually be proved before building anything.
Build the outcome.
We create the sourced data, tests, and workflow required to make the answer repeatable.
Keep it useful.
We stay with the work as sources and conditions change. Each outcome makes the next question easier. That ongoing involvement is the retainer.
The point is not a better demo. It is work your team can inspect, use, and build on.
What we are not
CMD+RVL is not:
- A data reseller
- Open-ended consulting or junior staff augmentation
We do not compete on how much data we have. Plenty of vendors do, and it is the wrong axis. We compete on what can be proved about the data you already own. You are buying ongoing senior delivery around one agreed outcome at a time.
Why teams call us
The usual shape is a number no one can trace and a decision that cannot wait. Who owns this bond? What changed since last quarter? What supports this exposure? Elsewhere, it may be a board question, a software renewal, or a model answer no one can verify.
Your team keeps the sourced data, tests, workflows, and decision records. The retainer keeps the work current and applies what was learned to the next valuable question.
Founder-led delivery
CMD+RVL is intentionally small at the point of responsibility. Drew and Zac stay close to scoping, architecture, QA, and handoff because the early shape of an outcome determines whether it can be defended later.
- Scoping
Founder review keeps the first target narrow enough to ship and specific enough to judge.
- Architecture
Production engineering and capital-markets experience shape what gets automated, cited, or left manual.
- QA
Receipts, assumptions, and escalation paths are checked before a result becomes part of a workflow.
The founders bring different operating instincts to the same problem: turn ambiguous data and AI work into a result that can be used, checked, and expanded deliberately.
Common questions
What does CMD+RVL do?
CMD+RVL is an applied AI firm for finance and procurement teams. We turn fragmented source records into finished work your team can verify and use. Every material number retains its source, as-of date, and path back to evidence.
Can I just ask Claude?
You can, and we do. The issue is that a model can give a traced number and an invented one with the same confidence, so we build the sourced data, tests, workflows, and decision records that make the answer inspectable before it becomes a decision.
What does the retainer include?
The retainer buys senior delivery around one agreed outcome at a time. We build against your existing stack and create the sourced data, tests, workflow, and decision records required to make the answer repeatable and keep it current.
What does our team keep?
Your team keeps the sourced data, tests, workflows, and decision records created along the way. What the team learns becomes the starting point for the next valuable question.
Do you need access to private data to start?
No. Our first proof case uses public SEC filings, so your team can inspect the same source documents we used. Start with one question that needs to hold up.
Our team
Drew Bittenbender
Co-Founder
Technology executive with 25 years building data products, analytics systems, and high-performance teams. Former CTO at Infinite Blue and VP Engineering at ColdLight (acquired by PTC for $105M). Holds a patent from Traffic.com (acquired by NAVTEQ for $179M).
Zac Ruiz
Co-Founder
Technology leader with 25+ years' experience, including a decade in securitization and capital markets. Former VP, Global Technology at DBRS, where he built transaction management systems and worked with quants across asset classes on data strategy.
Bring us one question that needs to hold up.