The five stages
DeconstructRedesignDesign phaseAugmentBuild
We map how the work gets done (Deconstruct)
Typically two to four weeks
What you get: a map of how the work runs today, with the time and cost of each bottleneck documented from your own records.
More about this stage
We start from your answers to the Intake Audit, and your staff map their own steps with a short guided form. Then we spend time with the people doing the job, in person or by video. We map what actually happens, where the time goes, where the errors come from, and where a system would help versus where it would only add something else to maintain.
The Intake Audit. After the first conversation and before any design work, you complete a structured Intake Audit form. It captures how the work actually flows, which documents and systems are involved, who touches what, where the time goes, and what happens when something goes wrong. Your answers are what the system design is built from. Skipping it means designing from assumptions, so we do not skip it.
Worth saying plainly: if we do not find enough here to justify the cost, we will say so, and you owe nothing further.
We fix the process, then design the system (Redesign)
Runs inside the design phase
What you get: the redesigned workflow, the documented savings you sign off on, and a design document with the written governance and a plan naming what we would build, what it would cost, and how long it would take. Nothing gets built until you approve it. If you stop here, the redesign and the design are yours to keep.
More about this stage
We remove steps that add nothing, standardize how the work is done, and decide where a person's judgment matters most. Only then do we design the system around the improved process. Before anything is built, you see its governance: what it can do, what it cannot do, who is allowed to use which parts, and what happens when it hits a question it should not answer.
We build it and try to break it (Augment)
Typically four to eight weeks
What you get: a working, tested, documented system. It does not go live until it passes.
More about this stage
We build it against your real work, connect it to the systems you already use where they allow it, and then deliberately try to make it fail. We feed it bad inputs, edge cases, and questions it should refuse. If it fails, we fix it and try again.
We train your people (Augment)
At launch and again after
What you get: trained people, written SOPs, and reference materials your firm keeps.
More about this stage
Owners and managers on one track. Staff on another. Then written procedures so it survives turnover.
We keep it running (Augment)
Tailored to each firm
What you get: reviews of how the system is performing and the adjustments they call for.
More about this stage
Your software changes, your work changes, and the rules around your work change. We recommend at least three months of monitoring after launch, with SageLogic available to adjust the system and correct issues as your team uses it. The length and scope of monitoring and ongoing support differ for every firm and are agreed with you in advance.
Systems thinking
We look beyond individual tasks.
A faster step adds little when the next one stays slow. Three ways we look at the whole system.
-
Symptom or cause
- Visible symptomStaff spend too long processing customer requests.
- Underlying problemsDisconnected systems, repeated data entry, unneeded approvals, and inconsistent procedures.
- Strategic responseRedesign the workflow before choosing automation tools.
-
Everything is connected
- Sales
- Operations
- Finance
- Customer service
A gap in one hand-off shows up as delays and errors further down the line.
-
Find the constraint
Speeding up any other step will not move more work through. Fixing the constraint can, although results vary by process.
Illustrations, not client results.
Inside the analysis
What we examine, how we think, what you receive.
Select a tab. This is the work behind the Deconstruct and Redesign stages.
What we examine
- 01
Business strategy
Revenue model, objectives, growth priorities, competitive position
Reveals Whether daily operations support the firm's goals
- 02
Operational processes
Workflows, procedures, handoffs, processing times
Reveals Unneeded work, delays, and bottlenecks
- 03
Coordination
Responsibilities, communication, dependencies between teams
Reveals Duplicated effort, unclear ownership, slow decisions
- 04
Information
Documents, data, stored knowledge, and how information moves
Reveals Scattered knowledge and information that is hard to find
- 05
Technology
Existing software, integrations, automation, digital tools
Reveals Systems that do not connect and tools that overlap
- 06
Financial performance
Labor cost, process cost, use of staff time, revenue opportunities
Reveals What each problem costs and what fixing it is worth
- 07
Risk and governance
Approvals, security, quality checks, compliance responsibilities
Reveals Weak controls and the oversight a system will need
One example: how the parts connect
- ObjectiveGrow revenue without adding administrative staff at the same rate.
- ConstraintProject managers spend too much of their week on documentation.
- ResponseRecover project manager capacity.
- RedesignOne standard way to collect, route, report, and approve.
- TechnologyStructured intake, document classification, AI-assisted reporting, project lookup.
- Expected outcomeLess routine work, clearer status, more capacity for billable work.
An illustration, not a client result. The depth of analysis matches the problem; not every engagement covers all seven.
Methods we use
Constraint analysisFind the bottleneck before investing in the solution.
Find what limits performance: where work piles up, where opportunities are lost, and which process caps capacity. Example: a firm has enough demand but cannot produce proposals fast enough, so the proposal process is redesigned before anyone suggests more marketing.
Goal factoringTurn business objectives into a practical plan of action.
Break a broad goal into the processes, decisions, and activities it depends on. Example: serving more clients may depend on faster onboarding, quicker document processing, less administrative work, and shorter approval cycles. Each one is analyzed and ranked.
Process mapping and workflow engineeringSee how work gets done, then design a better way.
Map the activities, handoffs, dependencies, and information flow behind the work. Example: client onboarding passes through sales, administration, accounting, and operations; duplicate data entry and unneeded approvals are found before a new workflow is designed.
Strategic technology alignmentChoose technology for the business need.
Check that a proposed tool solves the real problem and fits the systems you already have. Example: a firm asks for a customer chatbot, but replies are slow because staff cannot find accurate information internally, so internal knowledge retrieval comes first.
Economic impact and ROI modelingPrioritize improvements with measurable economic value.
Weigh build cost, running cost, recovered staff time, risk reduction, and revenue effects to estimate the return on investment (ROI). Example: an automation might save an estimated 600 staff hours a year; we check whether those hours become lower cost, avoided hiring, or more billable work.
Integrated systems architectureDesign systems that work together.
Design how people, information, software, and automated steps work together before anything is built. Example: separate systems for inquiries, documents, scheduling, and accounting get one plan that cuts duplicate entry and keeps permissions and controls in place.
What you receive
Deconstruct
- Business strategy reviewHow operational priorities line up with business objectives
- Operational discoveryA documented understanding of the main workflows
- Constraint analysisThe bottlenecks with the highest impact
- Process mappingA visual map of current activities and dependencies
- Technology assessmentA review of existing software and where AI could help
- Financial analysisThe documented cost of the problem and an estimate of costs and benefits
Redesign
- Goal factoringBusiness objectives broken into specific improvements
- Workflow redesignThe proposed future-state process
- Systems architectureA design for how software, data, and automation work together
- Implementation planningA prioritized roadmap with responsibilities and milestones
Augment
- AI implementationWorking, tested workflows and systems
- Performance evaluationTesting against the objectives set in design
- Knowledge transferDocumentation, training, and ownership of the system
Possible deliverables. Each engagement includes the ones its problem needs, agreed in writing before work begins. The analysis has value on its own, even when the right answer needs little or no AI development.
Technology selection
AI is one option among eight.
We start with the simplest answer that solves the problem. Select an option to see when it fits.
1. Eliminate the step
Eliminate the step
When it fits
The step adds nothing anyone uses.
Example
A second approval that always says yes.
2. Simplify the workflow
Simplify the workflow
When it fits
The work takes more steps or hand-offs than it needs.
Example
Five hand-offs reduced to two.
3. Standardize the procedure
Standardize the procedure
When it fits
Different people do the same task in different ways.
Example
One report template and one review checklist.
4. Improve existing software
Improve existing software
When it fits
A tool you already pay for can do it but is not set up to.
Example
Turning on a feature your practice software already has.
5. Integrate existing systems
Integrate existing systems
When it fits
The same information is typed into two systems.
Example
Connecting intake to the project system.
6. Conventional automation
Conventional automation
When it fits
The task follows the same fixed rules every time.
Example
Reminders, routing, and file naming.
7. AI assistance with human review
AI assistance with human review
When it fits
The task means reading or drafting from documents and language that fixed rules cannot handle.
Example
Drafting a report section from field data, checked by the project manager.
8. Custom software
Custom software
When it fits
No existing tool fits the way the work must run.
Example
A purpose-built review queue for one firm's approvals.
The work stays with your team. Editing and reviewing the output, signing the report, advising a client, approving a payment. Human judgment and oversight remain in every option above.
When options 1 to 6 solve the problem, we say so before anything is built.
Architect, then builder
We design the solution before we build the technology.
A contractor builds what someone specifies. An architect works out the requirements, the constraints, and the result first. SageLogic does both, in that order.
Some developers and consultants also work this way. The comparison describes two approaches, not specific companies.
What the finished system does
Every step, from request to release.
Every system we build follows one process, set by the governance you approve in design, from the moment a request arrives to the moment approved work goes out. The animation below runs through it once. Each step is explained underneath.
The steps are listed below.
The same process, in use
What one workflow looks like on screen.
This is the screen, or User Interface (UI), a team would use for a single workflow, here a monthly client report built from ten sources: four weekly field reports, a budget tracker, lab results, the schedule, client meeting notes, last month's report, and the report template. Add documents, tell it what you need, run it, and choose who reviews and who does the final edit. Nothing is released until both approve. Take the guided tour below, or explore it on your own.
Design preview with sample data. Files you add stay in your browser and are never uploaded or read. The project, client, figures, and staff names are invented, and the draft is simulated.
Want the parts in detail? See how the pieces of the system fit together.
What each step is
The same process, in words.
A request reaches the system, either because someone at your firm asked it something or because a document arrived. The system gathers what it needs from two places and nowhere else: the material your firm approved, and the systems you connected it to. It assembles a draft inside your own template, with a note of where each piece came from. Before that draft goes anywhere, it is tested against the limits you set during design. Anything outside those limits stops and goes to the person who owns that decision. Everything else waits at a single approval step until a named reviewer signs it off. Only then is the document released, and the whole run is logged with its sources attached.
Read each step in words
01 · In
A question or a task
The starting point. A person asks the system something, or a document lands in a folder or an inbox the system watches. Nothing runs until one of those two things happens.
02 · What it reads
Your approved sources
CITEDEverything the system reads is material your firm already owns. It reads it when a question is asked, brings back the passages it used, and names the file each answer came from, so your team can check it rather than trust it.
Your work product
- Final reports, technical memos, and letters
- Permit applications, submittals, and agency correspondence
- Proposals, scopes of work, and fee letters
- Executed contracts, engagement letters, and change orders
- Workpapers, calculation packages, and supporting schedules
- Field notes, logs, lab results, and project data tables
How your firm does the work
- SOPs, work instructions, and desk procedures
- QA and QC checklists and review criteria
- Report shells, letter templates, and standard language
- Internal standards, style guides, and naming conventions
- Onboarding and training material your firm already wrote
- Decision records, meeting notes, and project close-out summaries
- Knowledge captured from interviews with your senior staff during design
Reference material your firm holds
- Regulatory summaries and code interpretations your firm has written
- Agency guidance documents your firm relies on
- Client and matter records, where the engagement allows their reuse
- Subject files and research your team has already assembled
File types it can take
- PDF, including scanned pages read by character recognition
- Word, Excel, PowerPoint, plain text, and CSV
- Email messages and attachments, from mailboxes you name (must be edited for client privacy purposes)
- Photographs of documents, read by character recognition
- Records read from your connected systems, read only
Nothing outside this approved list is read. Not personal drives, not HR or payroll files unless you scope them in, and not the open internet.
03 · What it reaches
Your own systems
READ ONLYConnections to the software your firm already runs, so the system can look up a matter, a project, a client, or an invoice without anyone copying it across by hand. Common examples:
Document storage and email
- Microsoft 365: SharePoint, OneDrive, and Outlook
- Google Workspace: Drive and Gmail
- Box, Dropbox, and Egnyte
- NetDocuments and iManage, for law firms
Project, practice, or matter systems
- Clio, MyCase, and Filevine, for law firms
- CCH Axcess, Karbon, and TaxDome, for CPA firms
- Deltek Vantagepoint, BQE CORE, and Procore, for engineering and environmental firms
- dotloop and SkySlope, for real estate
Accounting and billing
- QuickBooks Online
- Xero
- Sage Intacct
- Oracle NetSuite
Client records and signatures
- HubSpot
- Salesforce
- DocuSign
- Adobe Acrobat Sign
These connections read. They do not write. Anything that would change a record in one of your systems requires your approval first, agreed during design and written into the build.
Which connections are possible depends on the product and the version your firm runs, and is confirmed during design. If your software is not listed here, that does not mean it cannot connect. Desktop software, such as installed tax or accounting programs, is handled case by case.
04 · Assemble
The draft
The system fills in your firm's template with what it found and notes the source of each piece. At this point it is only a draft. Nothing is sent, filed, or acted on until a person approves it.
05 · Check
Is it sure?
Before the draft goes any further, the system checks it against rules your firm writes down during design. There are two kinds of rule:
- Questions it may never answer on its own. For example, anything that calls for a professional opinion or a signature.
- Situations that always need a person. For example, a source is missing, two sources disagree, or a figure is above a limit you set.
If the draft passes every rule, it moves on to human approval. If it breaks any rule, it stops and goes to a named person instead.
05 · Escalation
Ask a named person
When the check fails, the system does not guess and does not carry on. The item stops and goes to the person at your firm who owns that decision, named in the system by role and by name. Work stops there until they answer.
06 · Gate
Human approval
Everything that passes the check waits here. A named reviewer at your firm reads the draft, changes what needs changing, and approves it. Nothing reaches a client, a regulator, or a court because the system decided it was ready.
07 · Out
Released and logged
The approved document goes out, and the run is recorded: what was asked, what the system read, what it produced, who approved it, and when. That record is what makes the work defensible later.
What it costs
Transparent, value-based pricing
Every engagement starts with the design phase at a flat fee. The build is priced from what the design documents, and that price is fixed in writing before any build work begins.
Step 1 (Deconstruct and Redesign)
Process and System Design
$1,500 flat fee
Covers the Deconstruct and Redesign stages: a map of how the work runs today, the redesigned process, the documented savings, integration planning, and the system design with its written governance.
Your design fee is fully credited toward implementation when you move forward with SageLogic.
Step 2 (Augment)
Build and Implementation
Starting at $3,500
The build is priced by complexity, technical requirements, integrations, and the business value documented in design.
The fee is based on the savings you signed off on during design, with a $3,500 minimum, and it is fixed in writing before the build begins.
After launch
Ongoing Support & Optimization
Tailored to each firm
After launch, SageLogic monitors the system and remains available to adjust it and correct issues as your team puts it to work.
Monitoring and ongoing support are tailored to each firm's system and needs, so the terms vary from one engagement to the next.
We recommend at least three months of monitoring and support after launch.
Savings and improvement figures are estimates, not a guarantee of results.
We document what the work costs you today
Hours per cycle, cycles per year, the loaded cost of those hours, and what a delay or an error costs when one happens. We build that figure from your own records where they exist, not from guesses, and you review and sign off on it in writing before the implementation fee is set.
No per-person and no per-use charges
Your bill does not go up because your team used it more than expected, and you never have to ask staff to ration it.
You keep what we build
The system, the documentation, and the training materials belong to your firm once the build is paid for. You are not renting it.
Built around your business
Every SageLogic system is designed around your redesigned process. You receive a clear scope and a fixed price in writing before any build work begins.