Define the question
Capture the customer segment, business problem, geography, constraints, and success criteria.
Evidence: a founder-approved brief
THE FOUNDATION · 01
Our business first.
A platform built to serve.
PROJECTBLUEPRINT
02 / MVP SCOPE
A managed AI operations platform: a subscription service combining dedicated AI agent teams and human expertise to help businesses design, run, and improve their operations.
Explore the first workflow ↗
First users: Smitty, Danny, and Chris.
Current stage: planning an internal web app.
01 / BUILD INTERNALLYResearch.
Start with context.
One workflow.
Research.
Make a decision.
Turn a business hypothesis into an evidence-backed market decision and a practical customer-validation plan.
Market Research MVP ↗02 / VALIDATE WITH CUSTOMERSStart with context.
Measure the outcome.
A founding cohort of 3–5 customers supplies required business context and SMART goals before the pilot begins.
Pilot requirements ↗03 / EXPAND WITH EVIDENCEOne workflow.
A reusable foundation.
Lead management follows market research. Broader operations, iOS, and voice remain later development.
Release sequence ↗The wider vision combines software subscriptions with research and advisory services. This site is the planning hub; the operational application is still to be built. The domain is struo.work; the final business name is undecided.
THE FIRST WORKFLOW · 02
From a hypothesis to an approved next step.
A mobile-friendly web app for the founders, focused first on market research for non-healthcare businesses. The initial customer segment remains open.
Gather traceable evidence
Research demand, alternatives, workflows, and adoption barriers. Link findings to sources and dates.
Evidence: sourced findings and visible gaps
Evaluate opportunities
Use explicit criteria and weights to compare segments. Distinguish observations, assumptions, and uncertain estimates.
Evidence: a transparent opportunity scorecard
Record the decision
Present the recommendation, uncertainty, and proposed action. A founder approves it or requests revision.
Evidence: a recorded decision and reviewer
Test with customers
Record interview findings, objections, and pilot interest. Update the hypothesis when evidence challenges it.
Evidence: a customer-validation record
Prepare lead management
Produce an approved customer profile and offer hypothesis for the next workflow.
Evidence: a usable, approved handoff
Proposed acceptance criteria
- All founders can sign in; retrieval respects their access permissions.
- Research answers cite supporting sources and acknowledge missing information.
- Actions that require approval remain blocked until approval is recorded.
- The team can trace recommendations, revisions, decisions, and outcomes.
- Completion time, corrections, failures, and usage costs can be measured.
THE ENGINE · 03
Trusted context. Controlled action.
01
Approved business context
Goals, documents, constraints, and authorized tool data
↓
02
Retrieve relevant knowledge
Permission-aware vector search and RAG, with source references
↓
03
Coordinate specialist agents
A conversational point of contact gathers research and drafts a plan
↓
04
Review, record, and improve
Human approval, outcome tracking, and approved customer-specific feedback
Security built into delivery.
Implement controls, operate them consistently, and verify them with recorded evidence and clear responsibilities.
Separate client data, access, and retrieval. Exact infrastructure boundaries still need a technical decision and isolation tests.
External outreach, spending, and changes to connected systems require explicit human approval. Research and drafting may be automated within approved scope.
Model a desired outcome, show assumptions, and propose a plan. Forecasts show ranges; they do not guarantee results.
Retain approved context and feedback, track results, and propose revisions. Operational changes remain subject to review.
Vector search, RAG, and learning—in plain language
Vector search finds material by meaning. Retrieval-augmented generation (RAG) supplies relevant source material before an AI answers. Initial learning means approved context and workflow improvements; unrestricted model retraining is outside the first release.
THE FOUNDING COHORT · 04
Context makes a pilot ready. Results establish value.
After the internal foundation works, test transferability with a founding cohort. A six-week pilot is proposed; dates, pricing, and success thresholds require agreement.
Required signup context
Collect the business need, current workflow, relevant approved information, SMART goals, baseline measures, constraints, and a named decision-maker.
Gate: complete, usable, authorized context
Agree the measurement plan
Set a scoped outcome, target, deadline, data source, review cadence, and approval owner. Track quality, time, corrections, failures, and service costs.
Gate: a baseline and agreed success criteria
Test value and continuation
Compare results with the baseline, document customer feedback, and assess paid continuation. Report uncertainty and other factors affecting results.
Gate: customer value and viable delivery economics
What counts as evidence?
SMART goals are specific, measurable, achievable, relevant, and time-bound. Signup data enables a relevant service; it does not establish willingness to pay. Internal use tests usability and reliability. Customer outcomes and paid continuation test whether the offer transfers. A change after the pilot does not by itself prove causation, and early qualitative findings should not be described as statistically significant.
THE PATH · 05
Build internally. Validate externally. Expand deliberately.
01 FIRST RELEASE
Internal research alpha
Build the mobile-friendly research workflow for the three founders, including knowledge retrieval, approvals, and outcome records.
Gate: usable workflow and verified controls02
Founding-customer pilot
Onboard 3–5 customers into dedicated environments with required context and agreed SMART goals.
Gate: measured value and paid continuation evidence03
Lead management
Reuse approved customer profiles and offer hypotheses. Extend the tested delivery pattern into follow-up workflows.
Gate: repeatable quality and viable costs04
Broader operations
Evaluate iOS and voice through internal testing. Expand into scheduling, business insights, and retention as evidence supports.
Gate: demand, capacity, and security readinessDELIVERY & RESOURCES · 06
Make ownership and cost visible.
At least 10 hours per founder weekly: 30 combined hours for planning, research, development, and testing. Operations are primarily in Indiana and remote; a separate legal entity has not yet been formed.
Market & workflow research
Lead market and legacy-system research. Proposed MVP focus: research briefs, customer-validation planning, and outcome measurement.
Detailed responsibilities need founder agreement
Technical & security planning
Lead technical requirements and security planning. Proposed MVP focus: retrieval, dedicated environments, access controls, and verification.
Technical boundaries remain to be specified
Agree delivery ownership
Confirm Chris’s workstream and assign accountability for onboarding, delivery, approvals, support, and acceptance testing.
All three founders are initial users and reviewers
$72–156 / month
Estimated monthly cash spending, including contingency, from the MVP planning summary.
$168–462 / month
Estimated monthly cash spending, including contingency, depending on cohort size and usage.
No ceiling approved
These are planning ranges, not current vendor quotes or approved spending. Founder time is separate. Confirm tools, usage assumptions, exclusions, and funding before committing.
THE NEXT DECISIONS · 07
The first workflow is set. Delivery details remain.
01
Customer segment & offer
Choose the first non-healthcare segment, scoped pilot offer, pricing, and acquisition approach.
Open02
Architecture & tools
Define dedicated-environment boundaries and select the minimum stack for retrieval, agents, security, and measurement.
Open03
Ownership & acceptance
Confirm founder responsibilities, approval owners, pilot targets, and release acceptance thresholds.
Open04
Budget & funding
Approve a monthly ceiling, validate cost assumptions, and determine the funding route.
Open05
Name & legal formation
Confirm the final business name and formation plan. struo.work is the current demonstration domain.
Open↗
The vision is a managed AI operations platform.
The next deliverable is a working market research MVP.