Build stage · 8 to 12 weeks · Starting at $60k
RoboCFO Pilot
Ship one production AI use case end-to-end with measured ROI.
Eight to twelve weeks. Charter, build, deploy, measure. One use case in production at the end, integrated with your finance systems, used by your team, with ROI captured against a documented baseline. Delivered by Glenn Hopper with an embedded team of finance technologists, data engineers, and AI specialists.
Best for
Who this is for
Finance teams ready to move past advisory and put AI in production. You have executive sponsorship, a use case in mind (or you've completed a Sprint that identified one), and you want a single capability live in a defined timeline with measurable outcomes.
If you're not sure which use case to start with, the Sprint is the right entry point. The Pilot picks up where the Sprint ends.
What's in the box
Seven concrete deliverables
The artifacts are real, the system is in production, the team is trained.
- Pilot charter document locking the use case, success criteria, scope, governance, and roles.
- ROI measurement framework with baseline metrics captured before build and target metrics defined for the readout.
- Production AI capability deployed in your environment, not a sandbox or proof-of-concept.
- Integration with finance systems as scoped (ERP, BI, data warehouse, planning tools).
- Documentation and runbook including operating procedures, troubleshooting, and ownership handoff.
- Training for finance team users so the people who'll use the system day-to-day are ready.
- Week-12 readout with measured outcomes against the baseline, recommendations, and decision frame for what's next.
How it runs
A four-phase structure
The build phase is where the embedded implementation team comes in.
- Phase 01
Weeks 1 to 2: Charter and scoping
Use case finalization, success criteria locked, integration planning, data access provisioning, technical architecture decisions, governance gates set. By end of week 2, every party knows what's being built and what success looks like.
- Phase 02
Weeks 3 to 6: Build phase
Embedded team executes. Model development or selection, integration work, data pipeline build or extension, security and access controls, internal testing. Glenn stays in the loop on architecture and product decisions; the build team handles execution.
- Phase 03
Weeks 7 to 9: Deploy phase
Production rollout, user onboarding, training delivery, monitoring setup, runbook documentation. Real users start using the system on real data. We're watching closely for issues and iterating fast.
- Phase 04
Weeks 10 to 12: Measure and refine
ROI capture against the baseline. Tuning based on real-world usage. Documentation finalization. Handoff prep. Week 12 readout with measured outcomes and a recommendation for what's next (typically Operations Standard for ongoing support, or a second Pilot for the next use case, or Transformation if multiple workstreams emerged).
What you bring
Real production engagement, real production inputs
- Executive sponsor at CFO or VP Finance level, committed across the full engagement.
- Use case alignment (one production use case, agreed upon at kickoff).
- Access to relevant systems (ERP, BI, data warehouse, planning tools as scoped).
- Data sample for development, with production data access established before build phase.
- IT and security collaboration for deployment, including the security review process.
- Internal user group (typically 5 to 15 finance team members) for training and feedback during deploy phase.
- Data and AI governance baseline or willingness to build it in parallel via the Governance Pack.
How we know it worked
Success criteria, locked at kickoff
Measured at week 12.
- Production system live in your environment, used by real finance team members on real data.
- ROI metrics measured against the documented baseline, with the result presented in the week-12 readout.
- Finance team users trained and operating the system independently.
- Documentation handoff complete including runbook, operating procedures, and ownership assignment.
- Charter success criteria all met validated explicitly by the executive sponsor.
A pilot that hits production but misses ROI is documented honestly. So is a pilot where the ROI exceeds expectations. The readout is straight reporting.
Pricing
Investment and what changes it
Starting at
$60k
Two tiers, scope-variable within each.
What changes the price
- Number of integrations and complexity (each major system integration adds scope)
- Team size and number of users to onboard
- Regulated-industry requirements (SOX, audit trail, model documentation, examiner review) typically adds 30 to 60 percent
- Whether you start from a recent Sprint or from scratch (Sprint compresses discovery)
- Custom data engineering work required if data is fragmented or undocumented
- Whether the model is built from scratch or fine-tuned from an existing platform
We scope and quote on the kickoff call, including which tier fits and where in the range you'll land. Published $60k floor is the entry point for Pilot Solo.
Two tiers
Pilot Solo
One use case, single-team rollout, standard integrations.
Best for finance teams of 5 to 30 deploying their first AI capability with one team owning the rollout. Typical scope: one use case (e.g., AP invoice processing, expense categorization, variance commentary generation, contract review), one team adopting it, integration with one or two existing finance systems.
Pilot Standard
One use case, multi-team rollout, regulated-industry scope, complex integrations.
Best for finance teams of 30 to 100 deploying a use case across multiple teams or sub-functions, or any team operating in a regulated environment (SOX, banking, insurance) where the audit trail and model documentation work meaningfully expands scope.
How we deliver
How we deliver
Glenn Hopper leads every Pilot directly through charter, scoping, deployment review, and the week-12 readout. The build phase is delivered by an embedded team of finance technologists, data engineers, and AI specialists.
The team brings implementation muscle to engagements that need to ship code, not slides. We do not sell platforms. Vendor selection in the Pilot is based on what fits your team and stack, not on referral relationships.
Where this fits on the spectrum
Before and after the Pilot
Starts with
RoboCFO Sprint in most cases. The Sprint identifies the use case and produces the directional ROI sizing the Pilot turns into a measured baseline. Buyers who already know their use case can skip the Sprint and enter the Pilot directly.
Leads to
Operations Standard in most cases (ongoing tuning, governance updates, use case expansion). Or a second Pilot for the next use case. Or Transformation if the engagement surfaces a multi-workstream program need.
FAQ
Common questions
How do we pick the right use case?
Either complete a Sprint first (which produces a directional recommendation), or bring a use case you've already aligned on internally. The Pilot kickoff converts your candidate into a charter with locked success criteria. If we don't think the use case will hit ROI, we'll say so during scoping rather than after.
Solo vs Standard, how do we choose?
Two questions decide it. First, will more than one team adopt the system in this engagement? If yes, you're in Standard. Second, are you in a SOX-regulated environment, banking, insurance, or a context with audit-trail, model-documentation, or examiner-review requirements? If yes, you're in Standard regardless of team size.
What happens if the pilot doesn't hit ROI?
We document honestly. The readout shows what was built, what was measured, and what the result was. Some pilots over-deliver on ROI and become flagship references. Some hit production but miss the projected ROI for reasons that surface during the engagement (data quality worse than expected, adoption slower than expected, market shift, etc.). In every case, you have a working production system, a documented baseline, and a clear next-step recommendation. We do not bill more if outcomes miss.
Can we run multiple use cases in parallel?
Not in a Pilot. Two or more concurrent production use cases is a Transformation. The Pilot exists to ship one capability well. If you need parallel workstreams, the right shape is the Transformation, which has the program structure to coordinate across them.
Who builds the actual code and infrastructure?
The embedded implementation team. Finance technologists, data engineers, and AI specialists who deliver production code in your environment. Glenn leads architecture and product decisions, the team handles execution. You can meet the team during scoping if useful.
What if we don't have a data warehouse?
The Pilot scope can include data infrastructure work, or the engagement can scope around what exists. Most Pilots involve some data engineering as part of the build phase. If your data is genuinely fragmented and undocumented, we'll surface that during scoping and decide whether to include foundation work in the Pilot or defer it to a separate engagement.
How does this compare to working with our existing IT or AI team?
The Pilot is not a replacement for your internal team; it's a force-multiplier. Most clients keep internal IT and AI staff involved throughout the engagement and use the Pilot as a way to ship a real capability faster while building internal capability through observation and structured handoff. The Operations Retainer that follows the Pilot is designed to keep your team in the lead with our team available for ongoing support.
What if the work surfaces something we should be doing instead?
That's a feature. The Pilot's first two weeks include charter validation. If we surface that the use case is wrong (better candidate found, market shift, internal priority change), we'll recommend re-charter or pause and document the finding. The cost of changing direction in week 2 is much lower than the cost of shipping the wrong thing in week 12.
Ready to put one capability in production?
Schedule a 30-minute kickoff call. We'll talk through your candidate use case, your team, and which Pilot tier fits. If the Pilot isn't the right shape after the call, we'll point you to the one that is.
Explore other engagements
Other stages on the spectrum
RoboCFO Sprint
$15k–$25k·3–4 weeks
Directional read on AI maturity and 90-day path
RoboCFO Governance Pack
$25k–$45k·4–6 weeks
Defensible AI policy, vendor review, rollout plan
RoboCFO Academy
Starting at $30k·8 weeks
Cohort program for finance teams, two tiers
RoboCFO Transformation
Starting at $500k·6–18 months
Multi-workstream AI program for finance at scale
Operations Retainer
Starting at $5k/month·Monthly recurring
Three tiers: Essentials, Standard, Embedded
Ask SAL