01Forecast
We forecast it.
How many contacts will arrive, and how long each will take. Built from history, calendars and campaigns, then checked against what really happened.
WFM – Beyond the SpreadsheetFictional data only
Beyond the spreadsheet.
AI in workforce management, shown working. For everyone in contact centres and BPOs: operations, training, quality, recruitment, leaders and the planners who tie it together. Interactive demos on fictional data, a practical library, and a community.
How workforce management works, in four steps
02Staff
We staff it.
Work out how many agents it takes to hit the service target, then allow for the time people are away. Erlang C is the classic start, and you can try it in the AI Lab.
03Schedule
We schedule it.
Place shifts, breaks and coaching so the right people are on the phones at the right half hour, fairly and within the rules.
04Run the day
We run the day.
Watch live demand, move breaks, swap tasks and call for help early. A small move at the right moment protects customers and people alike. Then we review, and the loop starts again.
Free to learnNo sign-up
No fees. No client data.
The AI Lab, the library and the team guides are free and run in your browser, so nothing you type is uploaded. Share what you learned using fictional or public examples, and leave what you hold at work.
- Staffing, shrinkage and forecast calculators
- Plain-English how-tos and a glossary
- Guides for every team, not only WFM
Show and tellInteractive
See it work.
One made-up day, one roster, one simple rule. Change something and watch cover, service level and breaks respond.
One made-up day, 22 agents
Raise unplanned absence and watch cover fall short. Then switch on smart break timing and see what moving breaks does to service level.
–Estimated service level (target 80%)
–Half hours short of agents
0Breaks moved by the rule
THE ROSTER, IN SPACE & TIME22 people. One connected day.
08:00 → 20:00 / TIMEDEPTH / AGENTS A1–A22
Drag to orbit · scroll over the scene to zoom · arrow keys / + / − when focused. On touch, drag sideways to orbit; scroll vertically to explore the page.
13:00–13:30
Calls 78Need 14Available 0Est. service level 0%
Swipe the chart sideways to see the whole day.
- On the phones
- Break or lunch
- Moved by the rule
- Fewer agents than needed
- Agents needed (dashed line)
Fictional day: 1,175 calls, 4-minute average handle time, 80% answered in 20 seconds. Real centres also have break rules, labour law and agent preferences that this demo leaves out.
Four places to start
Use the tools, read the guides, find your team, or ask people who have met the same problem.
AI Lab
Staffing maths, shrinkage, forecast accuracy and a prompt library for contact-centre work. Everything runs in your browser, so nothing you type is uploaded.
Open the labFor your team
How workforce management affects operations, training, quality, recruitment, leaders, systems and frontline agents, what WFM needs from you, and what to ask.
Find your teamLibrary
Plain-English how-tos, a glossary and templates built on fictional data, including a guide to using AI without leaking anything.
Browse the libraryCommunity
Ask a peer from any team, show what you built, swap what worked. Members sign in with LinkedIn or Google, and every new member is approved by hand.
See how it worksAI in WFM, the honest version
It saves real time on the right jobs. It also gets things wrong with confidence, so a person stays in the loop.
Good at, with you checking
- Drafting and summarising. Weekly ops notes, executive summaries, stakeholder emails.
- Explaining. What a formula does, why a forecast missed, what a metric means.
- Brainstorming. Scenarios, root-cause hypotheses, tests to run.
- Spreadsheet help. Writing and debugging formulas, cleaning messy columns.
- Sorting text. Grouping contact reasons or QA comments into themes.
Still needs a person
- Sign-off. Forecasts, headcount and schedules are decisions, not outputs.
- Your real numbers. A model knows nothing about your centre, and you must not paste confidential detail in to teach it.
- Arithmetic you cannot check. Do the maths in Excel or code, then ask AI to explain it.
- People decisions. Performance, pay and exits sit with HR and labour law.
- Anything confidential. If you would not post it on a public wall, do not paste it into a tool.
One rule keeps this place useful.
Nothing from work comes in.
People here run real operations, so the line is simple and absolute. Share what you learned, not what you hold. Swap real details for fictional ones before you post.
- Employer, client or customer names and identifiers
- Real agent names, IDs, schedules or performance
- Call recordings, transcripts or screenshots from internal tools
- Contract terms, SLAs, penalties, prices or revenue
- Anything marked confidential or covered by an NDA
- Personal information about anyone
The one ruleMembers area opens soon
Share what you learned. Not what you hold.
Nothing from work comes in: no employer, client or customer names, data, recordings or screenshots. Members will sign in with LinkedIn or Google and be approved by hand, and a new member's first posts are reviewed before they appear.