Approach
Understand the work before changing it.
Some engagements begin with an operational problem that needs investigation. Others begin with a clearly defined solution that needs to be designed, developed, or implemented. Our approach adapts to the project while maintaining the same focus on understanding requirements, delivering reliable solutions, and achieving practical results.
Start where you are.
You have a problem to investigate.
Work is slow, manual, or error-prone and the cause isn't clear. We start at Understand and Diagnose, then recommend and deliver the right fix.
Understand → Diagnose → Redesign → Implement → Measure
You have a solution to build.
You already know you need an application, integration, automation, AI tool, or dashboard. We confirm requirements and move straight into design and development.
Understand requirements → Implement → Measure
The five stages
- 1
Understand
Start with the goal, the stakeholders, and the constraints: budget, policy, systems, and time. Then learn how the work actually happens, which is often different from how the policy or the manual says it happens.
- Goals and success measures
- Stakeholder interviews
- Observation of the real workflow
- Existing forms, systems, and documents
- 2
Diagnose
Trace handoffs, wait states, exception loops, and duplicated effort back to their underlying causes. Measure a baseline wherever data allows, so improvement can be shown rather than asserted.
- Handoffs and wait states
- Exception and rework loops
- Duplicate entry and effort
- Baseline measures
- 3
Redesign
Develop future-state alternatives and choose for simplicity and operational reality. The best design isn't the most sophisticated one; it's the one your team can run on a busy Tuesday.
- Future-state alternatives
- Trade-offs made explicit
- Simplicity over sophistication
- Prioritized roadmap
- 4
Implement
Build and deliver the change. That can mean revising the process, configuring existing tools, connecting systems, automating workflows, implementing AI-assisted steps, or developing a custom application. Everything is tested, documented, and handed off.
- Process and policy changes
- System configuration
- Integrations and automation
- AI-assisted workflows
- Custom applications
- 5
Measure
Check adoption, throughput, errors, and effort against the baseline. Keep what works, adjust what doesn't, and decide together what to improve next.
- Adoption
- Throughput and cycle time
- Error and rework rates
- Next iteration
Illustrative example
What a redesign can look like.
A generic request-handling process before and after diagnosis. The fix here is mostly process and configuration: required fields, a routing rule, and automatic notification.
Illustrative only. Not a client process.
Current state · 7 steps
- Request emailed to a shared inbox
- Re-keyed into a tracking spreadsheet
- Waits for someone to triage itWait
- Missing information?Decision
- Sent back to requesterLoops back
- Waits for manager approvalWait
- Requester notified by hand
Future state · 4 steps
- Submitted through a form with required fields
- Routed by a written ruleAutomated
- Reviewed; only exceptions escalateDecision
- Requester notified automaticallyAutomated
Questions
Common questions.
Do I need to replace my software?
No. Replacement is sometimes the right call, but it's never the starting assumption. Many problems are solved by changing how existing systems are configured and used, or by changing the process around them.
Do I have to start with an assessment?
No. If you already know you need an application, integration, automation, or AI tool, we can start with requirements and move straight into design and development. An assessment is there for problems that need investigating first.
Do I need to know what the solution is before we talk?
No. Describe the symptoms: where things slow down, pile up, or get redone. Working out the cause and the right fix is part of the job.
Do you work remotely?
Often, yes. Most discovery and delivery work can be done remotely. Some engagements benefit from on-site observation, and that's settled during scoping.
Do you work across industries?
Yes. Workflow problems share common patterns across sectors. Where an industry has its own regulations or rules, those are verified with your stakeholders and subject-matter experts, not assumed.
How do you approach AI?
Practically. We look for specific tasks where AI reduces repetitive work or makes information easier to reach, such as classifying documents, extracting information, routing requests, or answering questions from internal knowledge. For sensitive workflows, solutions include access controls, human review, validation, and data handling that fits your confidentiality and policy requirements.
A problem to solve, or a project to build? Let’s talk.
Describe what’s getting in the way, or what you need built. You don’t need a finished spec to start.