Impact

Operational warehouse capacity management

Jadey turns shift planning and daily evaluation into a guided capacity run in the warehouse. Before the shift, Jadey identifies whether roles, stations and critical time windows are staffed in a sustainable way. After the shift, it checks whether the plan held up against operational reality: attendance, order situation, process data, throughput, warnings and data gaps are connected into a defensible daily finding. The effect does not come from another report, but from earlier decision capability. Critical undercoverage, short-notice replanning and overtime become visible earlier, before they become expensive in operations. At the same time, the findings from every run flow back into the next planning cycle. This makes warehouse control less reactive, more auditable and economically more robust.

-35.70%

critical capacity gap hours

-39.30%

overtime ratio in warehouse operations

+9.94%

throughput per paid labour hour

Challenge

Shift planning under daily reality

Warehouse planning is decided before the operating day. Whether the plan holds up only becomes visible in operation: order volumes shift, backlogs build up, cut-off times become tight, load peaks hit individual stations, employees are absent and role demand changes faster than a static shift plan can react.

At that point, the plan only shows how the day was intended, not whether operations remain deliverable under real load. For executive and logistics leadership, what matters is not the plan itself, but whether the right roles are available at the right time at the critical stations before throughput is missing, overtime builds up or orders can no longer be completed on time.

This is where operational damage starts: undercoverage becomes visible when performance is already missing. Overstaffing becomes visible when idle time has already been paid. Wrong role coverage becomes visible when a bottleneck blocks the next area. Without running feedback between plan, daily load and the next decision, warehouse control becomes reactive, with recurring bottlenecks, capacity bound to the wrong work and decisions that only emerge after operations have already paid for the mistake.

Solution

Operational capacity management in the warehouse

Jadey turns shift planning and daily evaluation into operational capacity management in the warehouse. The run connects demand picture, critical staffing windows, planning finding, running capacity warnings, live shift data, deviation explanation, approval and return into the next planning cycle. Executive and logistics leadership therefore receive more than a report: they receive a defensible decision basis for the active shift and the next planning period.

Before the shift, a planning finding is created for the upcoming planning period. It uses demand forecast, order situation, shift rules, existing Microsoft Teams Shifts data, open shifts, TimeCard or time-recording data where Teams Shifts / Time Clock is used in production, and findings from previous runs. Jadey does not change productive shifts without approval. Drafts for open shifts, person-related suggestions or productive Microsoft Graph effects require effect mode, purpose binding, approval and an audit trail.

During the shift, capacity checking continues. Jadey compares real attendance, role coverage, cut-off position, backlog and available process data with the planning finding. Critical deviations are carried as warnings, questions or escalation need so leadership and operations can act before the next decision point. Replanning remains human-approved; Jadey makes risk, cause and next decision visible.

After the operational shift, the daily evaluation is created. It connects target plan, time recording, warehouse and application logs, and available process data from connected systems into a finding on working time, attendance, process-related activity gaps, throughput, warnings and data gaps. Person-related indicators are not handled as rankings or automated performance judgements, but interpreted for a defined purpose with role, station, shift window, order mix and data quality.

The run does not end with a report. Jadey evaluates planning finding, running warnings, daily evaluation, operating blockers and approval need as one connected case. Verified outputs can feed KPI views; order situation, staffing situation, role coverage and warnings become input for later planning. The next planning cycle therefore does not start from zero again.

The strength lies in the learning loop. Every run checks whether the planning assumptions matched daily reality: which staffing windows were critical, which load peaks broke the plan, which data gaps prevented decisions and which approvals recurred. These findings do not remain a retrospective view; they flow traceably into the next planning finding.

Governance remains part of the operating model. Planning effects, person-related indicators and critical decisions are routed through approval, logged with purpose binding and returned to the responsible human owners. Jadey does not replace people management or HR decisions; Jadey makes operational workforce decisions more auditable, visible earlier and easier to connect to existing decision rights.

Metric	Before Jadey	After Jadey	Effect	Statement
Critical capacity gap hours	Index 100.00	Index 64.30	-35.70%	Fewer windows with missing role or station coverage under critical order load.
Reactive shift corrections	Index 100.00	Index 35.40	-64.60%	Fewer reactive reassignments, questions and manual interventions through earlier visibility before and during the shift.
Overtime ratio in warehouse operations	Index 100.00	Index 60.70	-39.30%	In the measured area, bottlenecks were identified earlier instead of being absorbed at short notice through extra work.
Throughput per paid labour hour	Index 100.00	Index 109.94	+9.94%	More operational output was created in the adjusted standard path from the same available working time.

The metrics refer to an eight-week measurement period with 112 evaluated shifts in a defined warehouse area. Teams Shifts planning, TimeCard or time-recording data where available, warehouse and app logs, process data and Jadey's planning, warning and daily-evaluation artefacts were evaluated. The index values show the observed relative change compared with the previous manual control baseline. Absolute workforce, order and location data remain anonymized for confidentiality reasons.

Critical capacity gap hours count windows in which role, station and order load together create a capacity risk. The throughput index was scoped against order mix, system incidents, training, manual corrections and missing mandatory data in the measured area. These special cases are carried as warnings and are not silently counted into the standard path.

The strongest operational change is the earlier visibility of critical capacity risks. Undercoverage, short-notice reassignment and overtime no longer appear only as a daily finding; they become a controllable decision case before and during the shift.

At the same time, Jadey connects planning, real attendance, order situation and process data into an economic view. The decisive question is not only whether a shift is staffed, but whether the available working time is translated into throughput with the right role and station mix.

A decidable run in this deployment means: mandatory data is present, data quality is sufficient, the planning finding, running warning or daily evaluation was created completely, warnings are visible, approval or escalation need is clearly named, mandatory evidence is present and no correction run for the same scope is required.

Technical depth

From forecast to the next planning cycle

The technical architecture separates schedule control, case guidance, planning finding, running capacity checks, daily evaluation, approval, KPI view and operating blockers. Jadey therefore does not guide the warehouse value stream as uncontrolled full automation, but as a bounded, auditable and documented operating process.

Recurring operations start through schedule control. In the standard run, the individual warehouse case then belongs to shift/ops. It is complete only when planning finding, running warnings and daily evaluation have been created, operating blockers are visible, approval or escalation need is documented and the return into the next planning cycle has been prepared.

Agents and roles in the deployment

Seven agent areas participate in the deployment. cron/ops ensures that planned runs start on time for shift windows, cut-offs and follow-up checks. shift/ops then holds outcome ownership for the warehouse case across planning finding, running capacity checks, daily evaluation, approval, escalation and return into the next planning cycle. shift/planner creates the planning finding for the upcoming planning period. shift/analytics evaluates actual data, working time, attendance, activity, process performance and warnings during the shift and creates the daily evaluation after the operational shift. shift/approvals reviews approvals for planning effects, person-related findings or critical replanning. shift/kpi aggregates verified outputs for KPI views. shift/admin resolves operating blockers when data access, Microsoft Graph, SQL or runtime are blocked.

Agent/role	Function in the deployment	Activity
`cron/ops`	Schedule control	Starts planned or event-adjacent runs for planning finding, daily evaluation and follow-up checks.
	`cron/ops` checks schedule, shift windows, cut-offs, locks, already active cases and catch-up needs. When a run is due, the agent hands a clearly scoped case assignment to `shift/ops`. Boundary: `cron/ops` holds no business outcome ownership, does not assess capacity and makes no approval or HR decision.
`shift/ops`	Outcome-owning case guidance	Keeps target, scope, status, next decision and closure of the warehouse case together.
	`shift/ops` takes over the case assignment from `cron/ops`, clarifies scope, date, location, team, planning period and effect mode. The agent commissions `shift/planner`, evaluates the planning finding and requests approval when planning effect or person-related relevance is present. During the shift, `shift/ops` commissions the running capacity check and clarifies critical replanning through the approval path. After the operational shift, `shift/ops` commissions the daily evaluation, evaluates warnings and operating blockers, and prepares the return into the next planning cycle. The specialised agents provide partial findings; `shift/ops` owns whether the case is closed, escalated or blocked. Boundary: `shift/ops` holds outcome ownership for the operational case, but does not replace business approval, does not make HR decisions and does not create productive Graph effects without defined effect mode and approval.
`shift/planner`	Planning suggestions and planning finding	Creates planning suggestions and planning findings for the upcoming planning period and produces planning artefacts.
	`shift/planner` uses demand forecast, order situation, shift rules, existing shifts, open shifts, historical daily evaluations and operating blockers. The agent creates the planning finding, plan JSON, Microsoft Graph payloads and, where applicable, drafts for open shifts. Boundary: `shift/planner` does not publish planning, delete shifts or assign people without approval and defined effect mode.
`shift/analytics`	Running capacity check, daily evaluation and evidence	Evaluates actual data, working time, attendance, activity, process performance and warnings, and creates daily-evaluation and warning artefacts.
	`shift/analytics` reads Teams Shifts, TimeCards, warehouse and app logs, and SQL process data. During the shift, the agent checks actual data, attendance, role coverage, activity, process performance, backlog, cut-offs, warnings and data gaps against the planning finding. After the shift, `shift/analytics` creates the daily evaluation as JSON and Markdown. Boundary: `shift/analytics` creates no write effect in planning, reassigns nobody, triggers no person-related measures and may not use raw findings without context as a formal assessment.
`shift/approvals`	Approval and governance gate	Reviews planning effect, person-related findings or critical decisions.
	`shift/approvals` returns decision-ready planning findings, critical replanning needs or person-related findings with operational relevance as approval, rejection, question or escalation. Human gates for warehouse/logistics leadership and HR/People remain part of the decision model. Boundary: `shift/approvals` does not replace technical remediation and creates no unregulated system effect.
`shift/kpi`	Management and KPI view	Consumes verified outputs from planning finding, daily evaluation, warnings and closing status.
	`shift/kpi` provides an aggregated view of on-time Jadey runs, critical staffing, decidable runs and further diagnostic metrics when planning finding, running warnings and daily evaluation are verified and usable for the defined purpose. Boundary: `shift/kpi` does not pass on raw data without purpose binding and does not replace operational or person-related decisions.
`shift/admin`	Resolve operating blockers	Handles auth, Graph, SQL, config, runtime or lock-related operating blockers.
	`shift/admin` checks technical error context, returns remediation, workaround, operating blocker status or escalation and keeps secrets out of repository and report artefacts. Boundary: `shift/admin` makes no business planning, throughput or HR decision.

Service hierarchy

flowchart TB
  cron_ops[cron/ops]
  shift_ops[shift/ops]
  shift_planner[shift/planner]
  shift_admin[shift/admin]
  shift_analytics[shift/analytics]
  shift_kpi[shift/kpi]
  shift_approvals[shift/approvals]
  microsoft_graph[(Microsoft Graph)]
  process_sql[(SQL process data)]
  wms[(Warehouse Management System)]
  artifacts[(Task artefacts / reports)]

  cron_ops --> shift_ops
  shift_ops --> shift_planner
  shift_ops --> shift_admin
  shift_ops --> shift_analytics
  shift_ops --> shift_approvals
  shift_planner --> microsoft_graph
  shift_planner --> process_sql
  shift_planner --> wms
  shift_planner --> artifacts
  shift_admin --> microsoft_graph
  shift_admin --> process_sql
  shift_analytics --> microsoft_graph
  shift_analytics --> process_sql
  shift_analytics --> wms
  shift_analytics --> artifacts
  shift_approvals --> artifacts
  shift_ops --> shift_kpi
  shift_kpi --> artifacts

Outcome ownership in the run

Phase	Responsible	Outcome
Trigger run	`cron/ops`	Due, unlocked case assignment.
Guide planning finding	`shift/ops`	Decidable planning finding with warnings and approval need; contributed by `shift/planner` and `shift/admin`.
Check running capacity	`shift/ops`	Warnings on cut-off risk, role coverage, station load and replanning need; contributed by `shift/analytics` and `shift/admin`.
Evaluate daily reality	`shift/ops`	Auditable daily evaluation with data quality and operating blockers; contributed by `shift/analytics` and `shift/admin`.
Clarify approval	`shift/ops`	Approval, rejection, question or escalation before planning effect, for critical replanning or after a person-related finding; contributed by `shift/approvals`.
Close outcome	`shift/ops`	Closing status and return into the next planning cycle; contributed by `shift/kpi` and `shift/admin`.
Check follow-up	`cron/ops`	Missing, blocked or late runs become visible; contributed by `shift/ops`.

Systems in the deployment

Jadey connects to the operational warehouse and Microsoft 365 system landscape. Microsoft Graph provides Teams shifts, open shifts and, where used, TimeCard or time-recording data. The Warehouse Management System provides order situation, station load and operating process states depending on integration depth. A connected SQL or reporting source provides throughput, process-related activity, reference data and data quality. Task artefacts and reports keep planning finding, running warnings, daily evaluation, approval, operating blockers and return into the next planning cycle traceable.

System	Use in the deployment	Description
`Microsoft Graph`	Shifts, open shifts, TimeCard or time-recording data and, where applicable, draft effects	Provides the shift, staffing-window, attendance and working-time view where Teams Shifts and Time Clock are used accordingly. Microsoft Graph is used for planning finding, running capacity checks, daily evaluation and, within approval boundaries, drafts for open shifts.
`Warehouse Management System`	Order situation, station load and operating process states depending on integration depth	Provides the operational view of orders, stations, backlogs, process status and critical load shifts in the warehouse.
`SQL process data`	Throughput, process-related activity, reference data and data quality	Provides structured process data with which throughput, activity gaps, order mix and data quality are contextualized for running warnings and daily evaluation.
`Task artefacts`	Evidence location for planning finding, running warnings, daily evaluation, approvals and closing status	Makes returns, approvals, operating blockers and reports auditable.

Artefacts in the deployment

The artefacts form the evidence chain from planning finding through running capacity warnings to daily evaluation and return into the next planning cycle. Every relevant step creates auditable evidence: case assignment, planning finding, planning artefact, capacity warning, daily evaluation, approval and operating blocker. Specialist agents create the partial artefacts; shift/ops holds outcome ownership for the case.

`case_assignment.json`

Owner: cron/ops

Meaning: Scopes timing, scope, location, team, planning period and trigger of the due run.

`planning_finding.md`

Owner: shift/planner

Meaning: Readable decision finding on planning need, critical staffing windows, role coverage, operating blockers and possible planning effect.

`plan.json`

Owner: shift/planner

Meaning: Machine-readable planning structure for audit, repeatability and return into later planning periods.

`graph_payload.json`

Owner: shift/planner

Meaning: Technical Microsoft Graph payloads for open-shift drafts or approved planning effect.

`capacity_warning.json`

Owner: shift/analytics

Meaning: Structured evidence on cut-off risk, backlog, role coverage, station load, data quality and required replanning during the shift.

`daily_evaluation.json`

Owner: shift/analytics

Meaning: Structured daily finding on working time, attendance, activity, throughput, warnings and data gaps.

`daily_evaluation.md`

Owner: shift/analytics

Meaning: Readable management and operations report that makes deviations, context and required action usable without raw-data analysis.

`approval_decision.json`

Owner: shift/approvals

Meaning: Documents approval, rejection, question or escalation for planning effect and person-related findings.

`blocker_status.json`

Owner: shift/admin

Meaning: Documents operating blocker, remediation, workaround or escalation.

From planning assignment to daily evaluation

The process diagram shows the standard run as a guided agent process. cron/ops starts the due shift run, hands it to shift/ops as an individual warehouse case, waits for the outcome and documents closure, follow-up or blockage. shift/ops keeps the case together and hands the business sub-assignments to shift/planner, shift/analytics, shift/approvals, shift/admin and shift/kpi.

flowchart TB
  subgraph approvals_lane["shift/approvals<br/>approval assignment"]
    ap1[Review planning effect, critical replanning<br/>or person-related finding]
    ap2[Return approval, question,<br/>rejection or escalation]
    ap1 --> ap2
  end

  subgraph admin_lane["shift/admin<br/>operating blocker"]
    ad1[Check Graph, SQL, WMS,<br/>configuration or runtime]
    ad2[Return remediation, workaround,<br/>blocker status or escalation]
    ad1 --> ad2
  end

  subgraph cron_ops["cron/ops<br/>shift-run assignment"]
    c1[cron/ops trigger]
    c2[Check shift window,<br/>cut-off or follow-up]
    c3[Sequence due<br/>warehouse cases]
    c4[Start shift/ops<br/>sub-assignment]
    c5[Wait for outcome, error,<br/>timeout or blocker]
    c6[Document run, follow-up status<br/>and blockages]
    c1 --> c2 --> c3 --> c4 --> c5 --> c6
  end

  subgraph shift_ops["shift/ops<br/>warehouse-case assignment"]
    s1[Start warehouse case:<br/>date, location, team, effect mode]
    s2[planner sub-assignment:<br/>request planning finding]
    s3[Wait for planning finding<br/>and planning artefacts]
    s4[Clarify approval need<br/>and planning effect]
    s5[analytics sub-assignment:<br/>run capacity checks]
    s6[Wait for capacity warning<br/>or question]
    s7[analytics sub-assignment:<br/>start daily evaluation]
    s8[Wait for daily evaluation<br/>and data-quality finding]
    s9[Evaluate findings, blockers<br/>and closing status]
    s10[kpi sub-assignment:<br/>aggregate verified outputs]
    s11[Return into next planning<br/>and closure to cron/ops]
    s1 --> s2 --> s3 --> s4 --> s5 --> s6 --> s7 --> s8 --> s9 --> s10 --> s11
  end

  subgraph planner_lane["shift/planner<br/>planning finding"]
    pl1[Check demand, order situation,<br/>shift rules and previous findings]
    pl2[Evaluate shifts, open shifts<br/>and staffing windows]
    pl3[Return planning finding, plan.json<br/>and Graph payloads]
    pl1 --> pl2 --> pl3
  end

  subgraph analytics_lane["shift/analytics<br/>capacity check and daily evaluation"]
    an1[Check actual data, attendance,<br/>backlog and cut-offs]
    an2[Return capacity warning<br/>or question]
    an3[Connect target plan, time data,<br/>WMS and process data]
    an4[Return daily evaluation<br/>as JSON and Markdown]
    an1 --> an2
    an3 --> an4
  end

  subgraph kpi_lane["shift/kpi<br/>KPI view"]
    kp1[Consume verified outputs<br/>from planning finding, warnings and evaluation]
    kp2[Provide KPIs, patterns<br/>and return signal]
    kp1 --> kp2
  end

  c4 --> s1
  s11 --> c5
  s2 --> pl1
  pl3 --> s3
  s4 -. approval assignment .-> ap1
  ap2 -. planning decision .-> s5
  s5 --> an1
  an2 --> s6
  s6 -. critical replanning .-> ap1
  ap2 -. run decision .-> s7
  s7 --> an3
  an4 --> s8
  s9 -. person-related finding<br/>or closing approval .-> ap1
  ap2 -. closing decision .-> s10
  s10 --> kp1
  kp2 --> s11
  an1 -. runtime or source blocked .-> ad1
  an3 -. evaluation blocked .-> ad1
  ad2 -. blocker status .-> s9

Implementation

This deployment can become a blueprint for your enterprise

The deployment shows how Jadey operates warehouse capacity as a guided value stream: from demand forecast and shift planning through operational execution to daily evaluation and the next planning decision.

What is implemented here for warehouse shifts can be transferred to other operational workforce processes where planning, actual data, throughput, approval and evaluation are currently separate: production, service teams, field operations or internal back-office units with recurring capacity windows.

In a first conversation, we define which operational capacity value stream is suitable for your enterprise, which data sources are available and which approvals remain human decisions. This shows where Jadey can quickly create more plannability, transparency and decidability.