A process-led field guide for internal forward deployed engineers. Map the operation, investigate the causes of lost value, size the opportunity and inspect how the work could change.
By the Forward Deployed Team · Updated 24 October 2026 · Illustrative scenarios; not customer results.
Start with the operating outcome. Follow the end-to-end process, identify a specific cause of lost value, then test the intervention. Eight process families below contain 24 working hypotheses; the visual investigations that follow illustrate selected mechanisms.
Demand, specification, purchasing, process effort and loss. Distinguish spend removed from capacity released.
03 / Capital resilience
Improve capital decisions
Asset use, investment discipline, liquidity and resilience. Preserve service and essential controls.
Evaluate every opportunity through: people and responsibilities · process and standard work · data and information · controls and decision rights · technology and adoption. AI is one possible enabler.
These are starting hypotheses, not diagnoses. Establish the baseline and the burden of proof before selecting a tool. All process families stay open as you scroll.
HR / 01
Workforce strategy and organization
01Demand
02capacity
03roles
04spans
05location
01.01
Role duplication may create avoidable coordination cost
01.02
Demand variability may drive overtime despite spare capacity
01.03
Unclear accountabilities may create repeated handoffs
Proof or rejection: what evidence is needed?
Map demand, roster capacity, work content and decision rights with managers and workers. Reject cuts that transfer work, breach coverage or worsen service.
Where AI could contribute
Aggregate workload and skill scenarios; organization design and staffing decisions require human judgment.
Unclear role requirements may cause repeated sourcing
02.02
Scheduling and approval queues may extend vacancy time
02.03
Channel spend may not produce qualified accepted hires
Proof or rejection: what evidence is needed?
Trace requisitions by role, candidate withdrawal, interview waits and accepted offers. Reject conclusions explained by skill scarcity or noncompetitive terms.
Where AI could contribute
Draft role evidence, coordinate scheduling and analyze channel outcomes; people assess candidates using relevant job criteria.
This sizes process effort, not candidate quality or vacancy revenue. Confirm realized effort reduction and hiring outcomes before valuing any other effect.
Generic induction may omit role-specific readiness tasks
Proof or rejection: what evidence is needed?
Join start dates, access tickets and readiness milestones; compare similar roles. Reject claimed gains that merely move work to managers.
Where AI could contribute
Assemble role-based checklists and flag missing dependencies; managers verify readiness.
Example sizing model / Time to productive readiness
Annual starts × verified readiness days recovered × value per productive day
Low / base / high illustration: 1,000 starts × 1 / 2 / 3 days × $200 = $200k / $400k / $600k potential annual productive capacity.
A day recovered is valuable only if useful work can be performed. Exclude mandatory induction and count manager or IT effort separately; this is not removed payroll.
Eligibility records may not reconcile to benefit invoices
05.03
Disconnected leaver records may sustain avoidable vendor charges
Proof or rejection: what evidence is needed?
Reconcile time edits, payroll exceptions and eligibility changes with approved records. Reject reductions that affect valid pay or benefit entitlements.
Where AI could contribute
Extract and reconcile exceptions; payroll and benefits owners approve corrections.
Size the opportunity / Annual processing capacity
Annual transactions × net minutes released ÷ 60 × loaded hourly cost
Recurring operational problems may drive avoidable attrition
07.02
Internal roles may stay vacant while relevant skills exist
07.03
Repeated unresolved cases may erode employee experience
Proof or rejection: what evidence is needed?
Use aggregated, access-controlled exit themes, mobility outcomes and case recurrence. Reject causal claims from correlations alone; test operational changes.
Where AI could contribute
Synthesize de-identified themes and support voluntary role discovery; no automated employment decisions.
Example sizing model / Avoidable replacement expenditure
The preventable shares are assumptions requiring a controlled test. Use verified recruiting, onboarding and contractor cash costs; avoid counting lost productivity twice.
Estimate released processing capacity from your own workload. This is an effort model, not a prediction of AI savings. Use the separate formulas above for cash, capital, margin, materials and energy.
Formula: tasks × minutes ÷ 60 × realization × hourly cost − running cost; subtract one-time cost in year one. Financial savings require an actual budget change. No inputs leave your browser.
Method and sizing notes
The opportunity structure combines three value lenses—cash release, cost optimization and capital resilience—with process, contract, working-capital, customer-management and investment analysis. Historical cost-transformation and finance operating-model references inform the process coverage. Every hypothesis requires evidence, a test that could reject it and accountable validation.
Historical transformation benefit ranges are not current benchmarks or AI effect sizes. This guide uses transparent low/base/high sensitivity examples instead. The eligible baseline, proposed change and realization route must be established for the organization. The examples are not additive, forecasts or customer results.
Prioritize by net value, strength of evidence, feasibility, time to value and service or control consequences. Deduct implementation, transition and recurring costs; assign one owner and one benefit record to prevent double counting.
01
HR / investigate the opportunity
HR service requests
The working hypothesis
Repeat policy questions and incomplete requests may absorb HR capacity and delay employee support.
What would establish or reject the case?
Tag an anonymized case sample; pilot source-linked answers and measure resolution accuracy, escalation and employee experience.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
02
HR / investigate the opportunity
Onboarding readiness
The working hypothesis
New starters may lose productive days because access, equipment and training requests are disconnected.
What would establish or reject the case?
Trace starter cohorts from offer to readiness; test a shared checklist and confirm access remains role appropriate.
Human action: Reconcile dependencies and notify accountable owners; approvals stay with system owners.
AI / analysis & workflow
Onboarding dependency review
OPERATING INPUT → REVIEW OUTPUT
Role-approved starter checklist and completion events
Access approval is pending; notify its owner and show the downstream readiness impact.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
03
HR / investigate the opportunity
Scheduling & overtime
The working hypothesis
Overtime may reflect skill coverage gaps and mismatched schedules rather than sustained understaffing.
What would establish or reject the case?
Replay demand by interval; include skills, breaks, availability and agreed constraints; pilot with employee input.
Human action: Review a feasible roster that respects availability and required competencies.
AI / analysis & workflow
Demand scenario planning
OPERATING INPUT → REVIEW OUTPUT
Demand intervals and skill coverage
Compare service coverage under demand scenarios before the scheduler reviews a roster.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
04
HR / investigate the opportunity
Skills & internal mobility
The working hypothesis
Vacancies may remain open while relevant internal skills are invisible across job families.
What would establish or reject the case?
Validate a skills taxonomy with employees and managers; evaluate match usefulness and missed matches before a pilot.
Measure the consequence
Internal fill rate; time to readiness; employee feedback
Enablers to investigate
Career interviews; opt-in skill profiles; role requirements; semantic matching
WORKED EXAMPLE 04Illustrative scenario
A project needs a skill already present elsewhere
ILLUSTRATIVE WORKING RECORD / HR-04
Requirement
Profile evidence
Gap
Process statistics
Prior project
Verify with employee
Equipment certification
Not recorded
Assessment needed
Availability
Opt-in profile
Manager discussion
ILLUSTRATIVE WORKING RECORD / HR-04
Requirement
Profile evidence
Gap
Process statistics
Prior project
Verify with employee
Equipment certification
Not recorded
Assessment needed
Availability
Opt-in profile
Manager discussion
Review interpretation Present explainable skills matches for employee and hiring-manager review.
Present explainable skills matches for employee and hiring-manager review.
Explore the applications
AI use cases: how could the work change?
Two visual examples · inputs, interpretation and human decisions
+
An illustrative AI application: inspect the source records, the interpretation and the human decision.
AI / records & interpretation
Skills & internal mobility: evidence assist
SOURCE RECORD → REVIEW NOTE
Skills match: quality analyst has process-statistics experience; certification still needs confirmation.
Link evidence · flag uncertainty
Inputs: Career interviews; opt-in skill profiles; role requirements; semantic matching.
Human action: Present explainable skills matches for employee and hiring-manager review.
AI / analysis & workflow
Role requirement extraction
OPERATING INPUT → REVIEW OUTPUT
Approved role descriptions and employee-confirmed skills
Separate required certification from a preference; expose any ambiguous requirement.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
05
HR / investigate the opportunity
Learning effectiveness
The working hypothesis
Completion counts may hide gaps between training and confident task execution.
What would establish or reject the case?
Compare observed task performance before and after training; control for prior experience and test retention.
Measure the consequence
Time to competency; task error; learner confidence
Human action: Use targeted practice and qualified observation; course completion is not certification.
AI / analysis & workflow
Training gap synthesis
OPERATING INPUT → REVIEW OUTPUT
Learner feedback and instructor observations
Group missed-setting checks into a practice topic; verify the theme with an instructor.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
06
HR / investigate the opportunity
Work design & handoffs
The working hypothesis
Unclear responsibilities may create repeated checks and interruptions across support teams.
What would establish or reject the case?
Map actual decisions through interviews and event samples; simplify one handoff and monitor control and service outcomes.
Measure the consequence
Handoffs; queue time; decision quality
Enablers to investigate
Role interviews; decision logs; process mining; workload sampling
WORKED EXAMPLE 06Illustrative scenario
One employee change passes through five teams
ILLUSTRATIVE WORKING RECORD / HR-06
Change step
Current owner
Repeated work
Prepare record
HR services
Data entered
Verify record
Local HR
Same fields re-entered
Approve change
Manager
Approval retained
ILLUSTRATIVE WORKING RECORD / HR-06
Change step
Current owner
Repeated work
Prepare record
HR services
Data entered
Verify record
Local HR
Same fields re-entered
Approve change
Manager
Approval retained
Review interpretation Clarify ownership and combine evidence collection before changing staffing assumptions.
Clarify ownership and combine evidence collection before changing staffing assumptions.
Explore the applications
AI use cases: how could the work change?
Two visual examples · inputs, interpretation and human decisions
+
An illustrative AI application: inspect the source records, the interpretation and the human decision.
AI / records & interpretation
Work design & handoffs: evidence assist
SOURCE RECORD → REVIEW NOTE
Interview synthesis: two roles both prepare the same change report; verify with the owners.
Link evidence · flag uncertainty
Inputs: Role interviews; decision logs; process mining; workload sampling.
Human action: Clarify ownership and combine evidence collection before changing staffing assumptions.
AI / analysis & workflow
Role clarity synthesis
OPERATING INPUT → REVIEW OUTPUT
Interview notes and documented decision rights
Two teams describe the same check; ask owners whether one evidence collection can serve both.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
07
HR / investigate the opportunity
Payroll exceptions
The working hypothesis
Late time corrections and inconsistent rules may generate avoidable payroll rework.
What would establish or reject the case?
Trace exceptions to source; test detection on historical pay runs and manually review both missed and false flags.
Measure the consequence
Correction rate; late adjustments; pay accuracy
Enablers to investigate
Payroll interviews; time records; pay rules; anomaly detection
WORKED EXAMPLE 07Illustrative scenario
A missing shift code becomes an off-cycle payment
ILLUSTRATIVE WORKING RECORD / HR-07
Record
Signal
Review
Shift 301
12 hours
Premium code absent
Shift 302
8 hours
Standard code present
Correction
Pending
Employee confirmation
ILLUSTRATIVE WORKING RECORD / HR-07
Record
Signal
Review
Shift 301
12 hours
Premium code absent
Shift 302
8 hours
Standard code present
Correction
Pending
Employee confirmation
Review interpretation Ask for correction before cut-off; payroll reviews changes before payment.
Ask for correction before cut-off; payroll reviews changes before payment.
Explore the applications
AI use cases: how could the work change?
Two visual examples · inputs, interpretation and human decisions
+
An illustrative AI application: inspect the source records, the interpretation and the human decision.
AI / records & interpretation
Payroll exceptions: evidence assist
SOURCE RECORD → REVIEW NOTE
Exception flag: 12-hour shift has no applicable premium code; check the underlying agreement.
Link evidence · flag uncertainty
Inputs: Payroll interviews; time records; pay rules; anomaly detection.
Human action: Ask for correction before cut-off; payroll reviews changes before payment.
AI / analysis & workflow
Pay-rule explanation
OPERATING INPUT → REVIEW OUTPUT
Effective pay rules and validated exception records
Draft why the premium may apply and link the applicable rule for payroll review.
Validate: Compare the draft interpretation with a verified sample and record missing or incorrect conclusions before changing the process.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
08
HR / investigate the opportunity
Recruitment queues and role clarity
The working hypothesis
Unclear requirements and delayed interview decisions may extend vacancy time even when suitable candidates are available.
What would establish or reject the case?
Trace role approval, sourcing, interviews and offer decisions by comparable role. Reject the hypothesis if skill scarcity or offer terms explain the delay.
Measure the consequence
Vacancy days by stage; candidate withdrawal; accepted offers; hiring-team effort
Enablers to investigate
Recruiter and manager interviews; requisition milestones; scheduling records; consent-aware process analysis
WORKED EXAMPLE 08Illustrative scenario
A candidate waits while interview decisions queue
ILLUSTRATIVE RECRUITING JOURNEY
Stage
Observed time
Evidence
Role approved
Day 1
Requirements signed off
Interview completed
Day 12
Relevant experience discussed
Panel decision
Day 21
9-day approval queue
ILLUSTRATIVE RECRUITING JOURNEY
Stage
Observed time
Evidence
Role approved
Day 1
Requirements signed off
Interview completed
Day 12
Relevant experience discussed
Panel decision
Day 21
9-day approval queue
Review interpretationInvestigate the nine-day decision queue. Agree role criteria and decision ownership before sourcing more candidates. No candidate is scored or selected by this example.
Separate the approval queue from sourcing time; test a time-boxed human review and check candidate experience and quality of hire.
Explore the applications
AI use cases: how could the work change?
Two visual examples · inputs, interpretation and human decisions
+
An illustrative AI application: inspect the source records, the interpretation and the human decision.
AI / process evidence
Requisition and scheduling timeline
EVENTS → WAITING TIME
Interview day 12 → decision day 21
Flag an approval queue for the hiring manager
Inputs: Approved role requirements, process timestamps and interviewer availability.
Human action: Clarify decision ownership and review candidates against relevant job criteria.
AI / preparation
Role-criteria evidence pack
APPROVED REQUIREMENTS → HUMAN REVIEW
Summarize the agreed tasks and open evidence questions.
Link each requirement to its approved source
Validate: Recruiter and manager check accuracy and job relevance. The tool does not decide employment outcomes.
What would weaken the hypothesis?
The proposed change fails the measured outcome, shifts work elsewhere, loses required controls or costs more than the verified benefit. Investigate those results before scaling.
Questions & answers
Before testing an opportunity
Where can AI help in hr?
Potential areas include hr service requests, onboarding readiness, scheduling & overtime, skills & internal mobility, learning effectiveness, work design & handoffs, payroll exceptions, recruitment queues and role clarity. Each is a hypothesis to test against operating evidence.
Are these savings or measured results?
No. All figures, records and overlays are illustrative scenarios. Validate the baseline, net cost, feasibility and service outcomes before making a business case.
Which enablers are needed?
Start with process interviews, reliable source records and a measurable baseline. Depending on the task, useful enablers include document extraction, retrieval, classification, forecasting and workflow integration. Assign a person to review exceptions.
What should a first pilot establish?
A pilot should establish whether a specific intervention changes the measured outcome against a credible baseline, how often it fails, who reviews exceptions and what it costs to operate.
