01
The workflow is expensive or slow
Repeated handling, manual transfer, rework, or delayed decisions are visible, but the root constraint is not yet agreed.
Recommended first engagementFixed discovery scope
Choose the workflow worth building before choosing the AI.
Map the current work, compare the right intervention, define data and action boundaries, and leave with a pilot scope your team can evaluate before a larger commitment.
A small, paid first step
One workflow.
A decision you can use.
$300 USD / fixed scope
The Starter Blueprint is a focused assessment, not the full enterprise discovery package. You keep the written findings and do not have to buy implementation.
Any applicable taxes are shown separately in the quote. Scope, availability, payment terms, and dates are confirmed before you commit. Additional workflows or deeper investigation need a separate quote.
- Scope boundary
- One workflow, up to two existing systems, and up to two client participants.
- Working sessions
- One 60-minute remote workshop and one 30-minute findings walkthrough.
- Your preparation
- Allow about 30 minutes to describe the workflow and prepare up to five redacted examples, totaling no more than ten pages.
- Delivery target
- Target: 3 to 5 business days after the workshop and receipt of the agreed inputs. The start date and delivery date are confirmed in the written scope.
Your written decision brief
- A current-workflow map with the bottleneck, owner, exceptions, and unverified baseline recorded.
- A comparison of up to three options: improve the process, automate it, or test AI.
- A proposed system sketch, human-review boundaries, and unresolved integration questions.
- A first-step scope, draft acceptance checklist, exclusions, and a build, experiment, defer, or no-build recommendation.
Not included: No code, prototype, live-system access, data cleaning, model benchmark, security/compliance audit, or implementation is included. One consolidated clarification round is included; new requirements are separately scoped.
When to start here
Use discovery to remove a specific investment uncertainty.
02
There are too many possible AI ideas
Teams have a list of agents, copilots, automations, or vendor options without a defensible order of investment.
03
Data and permission boundaries are unclear
The opportunity looks useful, but source quality, access, actions, approvals, and retention have not been mapped.
04
A pilot exists but production is uncertain
The demonstration works, while quality, integration, exception handling, operating cost, and ownership remain unresolved.
What you receive
Four deliverables that turn discussion into a build decision.
The package states what is known, what remains an assumption, what is excluded, and which evidence is required before production.
01Decision artifact
Current workflow and outcome baseline
People, triggers, records, systems, decisions, handling effort, failure points, exceptions, owners, and the result that should change.
Decision unlocked
Confirms whether the problem is valuable and bounded enough to pursue.
02Decision artifact
Prioritized opportunity map
Candidate AI and automation interventions compared by business value, feasibility, data readiness, operating risk, and change effort.
Decision unlocked
Shows what to test first and what should remain manual or deferred.
03Decision artifact
Target system and control boundaries
Proposed interfaces, integrations, context sources, model or rule responsibilities, actions, approvals, monitoring, and ownership.
Decision unlocked
Makes the proposed system understandable before code or vendor commitment.
04Decision artifact
Pilot scope and acceptance evidence
Representative cases, success and refusal expectations, integration assumptions, latency and cost checks, exception paths, exclusions, and rollout conditions.
Decision unlocked
Creates a buildable pilot brief and a clear production decision gate.
Example blueprint view
The system is mapped around the work and its boundaries.
Illustrative workflow map
Not a generic architecture templateTrigger
Work enters
Context
Sources allowed
Decision
Rules and AI
Action
Tools bounded
Owner
Review and recovery
Protected by
Permissions, validation, approval, and refusal behavior
Measured by
Completion, quality, exception rate, latency, and cost
Operated with
Monitoring, replay, escalation, runbooks, and ownership
How the engagement works
Discovery ends with a decision, not an inspiration deck.
- STEP 01
Prepare the operating context
Review existing notes, sample records, current systems, stakeholders, constraints, and any attempted solution before the working session.
- STEP 02
Map the real workflow
Follow the work from trigger to completion, including exceptions, judgment calls, access boundaries, and manual recovery paths.
- STEP 03
Compare intervention options
Test whether process redesign, deterministic automation, retrieval, an AI copilot, an agent, product changes, or no build is the most defensible move.
- STEP 04
Deliver the decision package
Turn findings into a workflow map, opportunity order, target architecture, pilot definition, acceptance evidence, assumptions, exclusions, and next-step recommendation.
Fit and boundary
Useful when the decision matters. Unnecessary when it does not.
Good fit
- A repeated workflow has visible cost, delay, rework, or decision pressure
- Several AI or automation opportunities need a defensible order
- A pilot needs production criteria and operating ownership
- An architecture or vendor commitment has material switching cost
- Internal stakeholders need one shared definition of the workflow
Probably not needed
- The requirement, acceptance criteria, integration, and owner are already agreed
- The task is a small deterministic integration with little uncertainty
- The goal is only a general AI presentation or trend workshop
- No stakeholder can provide representative examples or operating context
- The organization needs temporary staffing rather than a workflow decision
Written boundary
Workflows, stakeholders, systems, artifacts, sessions, assumptions, and exclusions are agreed before work starts.
Separated costs
Discovery cost, implementation cost, and recurring model or infrastructure cost remain separate decisions.
Evidence before expansion
A larger rollout follows only when the pilot's quality, operation, and business evidence support it.
Questions before discovery
Know what this engagement does and does not do.
What does the $300 Starter Blueprint include?
One workflow, up to two existing systems, and up to two client participants. One 60-minute remote workshop and one 30-minute findings walkthrough. You receive a workflow map, option comparison, system sketch, and draft acceptance checklist. No code, prototype, live-system access, data cleaning, model benchmark, security/compliance audit, or implementation is included. One consolidated clarification round is included; new requirements are separately scoped.
How long does the Starter Blueprint take?
Target: 3 to 5 business days after the workshop and receipt of the agreed inputs. The start date and delivery date are confirmed in the written scope. Allow about 30 minutes to describe the workflow and prepare up to five redacted examples, totaling no more than ten pages. Client preparation, the workshop, and the findings walkthrough total about two hours per participant, excluding the clarification round. Missing inputs or expanded scope can change the schedule.
Is the Workflow Blueprint only for enterprise teams?
No. It is useful for an enterprise business unit, a product team, or a founder when the workflow, data, integrations, and operating outcome need to be clarified before a meaningful build estimate or platform decision.
Does the Blueprint include software development?
The core engagement produces a decision-ready scope rather than a production build. A small technical spike may be recommended when one uncertain integration, data source, or model behavior must be tested before the pilot can be scoped responsibly.
Will you recommend a specific AI platform or model?
Only when the workflow and constraints justify it. The comparison can include deterministic automation, existing software, custom code, retrieval, model providers, agent frameworks, or keeping part of the work manual. The workflow comes before the tool.
What information is useful before the engagement?
A short workflow description, sample inputs and outputs with sensitive data removed where necessary, current tools, known delays or rework, responsible roles, relevant policies, and any prior pilot findings are useful. Perfect documentation is not required.
Can our internal team implement the resulting roadmap?
Yes. The deliverables are designed to support an internal team, another approved implementation partner, or a separately scoped ZamDev build. The workflow, decisions, assumptions, and acceptance evidence remain available to the organization.
How is the commercial scope defined?
The number of workflows, stakeholders, systems, sample records, workshops, and technical uncertainties are agreed before work starts. The written scope also states exclusions and whether any technical spike is included.
[ Start with the decision ]
Choose one workflow worth understanding properly.
Share the current process, systems, stakeholders, constraints, and outcome. You will receive a direct reply on whether a Blueprint, technical spike, pilot, or simpler next step fits.