Playbook

A second screen that helps UK contactcentre agents answer faster and stay well.It runs in a browser tab beside your phone system.

Agent assist platform
For UK contact centres
Feature-complete

Talk to me about Playbook

← All products

The problem

The answer hunt, while the customer waits.

Agents dig through shared drives, intranet pages and old emails with a customer on the line. Managers have little time to coach, and the people on the phones burn out.

Firms regulated by the FCA also have to show evidence of good customer outcomes under Consumer Duty, and that evidence usually lives in a spreadsheet.

What it replaces

Policy documents on SharePoint and shared drives
Wiki and intranet searches
Call-outcome spreadsheets
Whiteboard leaderboards and sticky notes

It runs next to whatever phone system you already use, so there is no integration project, and agents get answers from your own knowledge without being watched.

How it works

Set up by me. Used in a browser tab.

01
Workspace
I set up your workspace and your first admin. There is no self-serve signup.
02
Your setup
Your admin adds agents and teams, loads your articles and documents, and sets up queues, call outcomes, reason codes and disclosures.
03
On the calls
Agents open Playbook in a tab beside the phone system, to find answers, get AI help and log each call.
04
On the floor
Managers coach in real time, see team wellbeing and performance, and export Consumer Duty evidence.

Nothing to install and no IT project. You provide your first admin, your agent list, your knowledge, and how your queues work.

What it does

Everything the floor needs, in one tab.

Answers during the call

Plain-language help drawn from your own knowledge, while the customer is still on the line. The assistants run on Anthropic’s Claude Sonnet 5.

Support BotTier 2+Answers agent questions in plain language from your own knowledge.
Search by meaningTier 3+Finds the right document even when the words don’t match.
Call ScribeHelps agents wrap up a call quickly once it ends.
Objection CoachPhrasing for a customer objection, plus two or three probing questions.
Knowledge HubYour articles, searchable, with images and ratings.
Call loggingOutcome, reason code, vulnerability indicators and disclosure checks on every call. AI call summaries from Tier 2.
Callback calendarAgents schedule callbacks and get a reminder when one is due.

Why it holds

Every AI answer carries a fixed Consumer Duty safety preamble that no workspace can edit, and each agent’s AI use is capped on the server before any request is made.

PlaybookAssistSaleCallback

Agent

cooling-off period for a mid-term cancellation?

Playbook

Checking your knowledge

Customers have a 14-day cooling-off window from the start date.

Within it, confirm the date and process a full refund.

Agent

Policy sale — Personal Injury (£420/yr)

Sale logged

Level 6Level 7+60 XP

Callbacks this week

Reminder 14:15

MonTueTodayThuFri

10:00D. Moss

12:30R. HaleL. Grant

14:30K. AhmedJ. Patel

You have a callback in 15 minsK. Ahmed · Today 14:30

From your own knowledgeBrowser tab · zero integration

Outcome loggedXP and level-ups

Callback calendarReminded when it is due

Coaching, not surveillance

Managers help agents in the moment and see how the team is doing, without watching any one person.

Whisper coachingA private on-screen tip to an agent during a call.
Anonymous team moodManagers see the team picture, never one person’s mood.
Mood check-inAgents record how they feel at the start of a shift and see their own trend.
Abusive-call alertsManagers are alerted the moment an agent flags an abusive call.
Call reviewManagers review logged calls and add coaching notes.
Check-in nudgePrompt an agent to check in, at most once a day across all managers.

Why it holds

Whisper tips are never saved anywhere. Team mood stays hidden until at least 75% of the team has answered and at least three people have, and a “someone may be struggling” signal names nobody.

PlaybookActivityCallsAlerts

Team now

3 online4 online

J. OkaforRetentionOnline

S. PatelSalesOnline

L. TorresRetentionAwayOnline

R. HughesSalesOnline

M. BrennanRetentionOffline

AllSaleCallbackNo Sale

J. OkaforSale

S. PatelRenewed

R. HughesCallback Set

Callback Thu 3pm · annual plan

L. TorresNo Sale

Abusive alerts

2 unacknowledged3 unacknowledged

R. HughesRetention

Severe

M. BrennanSales

Standard

L. TorresRetention

Standard

Team activitySelf-reported presence

Call logExport as CSV

Abusive alertsSelf-reported by agents

Motivation built in

The parts of the shift that keep people coming back.

XP, missions and rewardsLevels, coins, missions and achievements, plus a shop for avatars, emblems and app themes.
Training gamesQuizzes, flash cards and more, built from your own policies. AI can generate them from a document from Tier 2.
Coin ArcadeHead-to-head mini-games between agents.
LeaderboardsLive team activity, leaderboards and team missions.
CelebrationsConfetti when a call is logged, and a moment for every level-up.
AnnouncementsManagers post updates to the team feed.

Why it holds

Arcade results are provably fair: the server publishes a fingerprint of the random seed before the game and reveals the seed after it.

PlaybookMissionsArcadeShop

Missions · this week

1,120 coins

Log 200 outcomes164/200

+80

Five-day check-in streak3/5

+50

Finish the vulnerability quiz4/55/5

+120

Team: 50 retention saves38/50

+200

Coin Arcade · coin flip

25 coin stake

Seed fingerprint9f2c · e71a · 41abSeed checked

M. BrennanHeads+25 coins

L. TorresTails

Heads

Shop

1,240 coins940 coins

AvatarPilot200 coins

EmblemStarburst300 coinsEquipped

ThemeMidnight500 coins

Missions and rewardsCoins for the shop

Coin ArcadeProvably fair

Avatars, emblems, themesBought with earned coins

Consumer Duty evidence

Outcomes, vulnerability and disclosures, organised as you work rather than rebuilt in a spreadsheet at the end of the month.

Control roomOutcomes, vulnerability and disclosures, by queue and by brand.
Report and CSV exportA written report and a CSV, ready for whoever asks.
Queue setupEach queue has its own outcomes, reason codes and disclosures.

Why it holds

Any figure backed by fewer than five agents is hidden, trends included, and it can’t be worked out by subtracting the others. The written report is produced without AI: it organises, and never infers.

PlaybookOutcomesVulnerabilityReport

Good outcomes · by queue

Privacy kept

Retention82%

Sales74%

Collections79%

ComplaintsFewer than 5 agents · hidden

Last 30 days

38Vulnerable

97%Disclosed

4Queues

Vulnerability flagged

Retention18

Collections11

Sales9

Consumer Duty report

CSV ready

1. Outcomes

2. Vulnerability

3. Disclosures

report.csv

Small groups suppressedBy queue and brand

VulnerabilityDisclosures checked

Written without AICSV export

Knowledge that keeps up

Your admin shapes the knowledge and the limits, without a development team.

Knowledge Gap RadarTier 3+Groups the searches your knowledge answers badly, labels each group, and drafts the missing article.
Articles with historyTier 2+Rich-text articles and documents, with earlier versions kept.
AI summary templatesTier 3+A summary template per queue.
AI usage capsTier 2+Daily and weekly AI limits per agent.
Multiple brandsTier 4Several brands and their queues in one workspace.

Why it holds

In Knowledge Gap Radar the AI only names each group of searches. It never counts or ranks them, so the numbers you act on are real.

PlaybookQueuesAI limitsAnalytics

Queue setup

Saved

QueueOutcomesDisclosuresAI summary

Retention63On

Sales52On

Complaints44Off

Collections53OffOn+ New queue

AI token limits

Daily50,00060,000

Weekly280,000

Agent overrides

Saved

JOJ. Okafor60k / day

SPS. Patel80k / day

RHR. Hughes50k / day

Last 30 days

1,284Total calls

76%Resolved

12%Escalated

By team

Retention612 calls

Sales449 calls

Complaints223 calls

Outcomes, disclosures, AI summariesSet per queue

Daily and weekly capsPer-agent overrides

AnalyticsExport as CSV

Who it's for

Built for the people on the phones.

A good fit

UK contact centres, from five agents up.
Especially FCA-regulated teams in consumer credit, insurance and debt advice, where Consumer Duty evidence matters most.
Bought by operations directors, championed by contact centre managers, and used every day by agents.

Not the right fit

Centres that want call recording, QA scoring or agent monitoring.
Very large estates that expect deep integrations.

What it deliberately doesn’t do

No telephonyNo dialler, softphone, call recording or CRM, so there is nothing to integrate.
No scoringNo AI scoring of agents and no keystroke monitoring. Scoring penalises the off-script care vulnerable customers need.
No individual moodManagers only ever see the team picture.

Founding programme

Be one of ten.

I’m taking up to 10 founding customers over the first 18 months.

In return: case-study rights, logo use, and a feedback call each quarter. Tier 2 and above, chosen at my discretion. If you want one of the ten, make your case.

Security and trust

Built to be trusted with a contact centre.

Your data, walled off
Every customer table is protected by row-level security in the database. A request that doesn’t name a workspace sees nothing.
EU hosting
Hosted in the EU, with files held in an EU jurisdiction. AI answers are processed by Anthropic under standard UK and EU transfer safeguards.
Sign-in
Secure sign-in with four roles: agent, manager, admin and Apexify operator. Sign-in tokens live in memory, never in browser storage.
An audit trail nobody can edit
Operator actions are written to tables the database won’t let anyone change or delete.
Erasure
Erase a whole workspace, audited before it runs, or a single agent’s data from Tier 2.
Accessible to agents
High contrast, reduced motion, a dyslexia-friendly font and zoom. Button text on your brand colours is held to accessible contrast.

Does it connect to our phone system?

No. It runs in a browser tab next to whatever you use now, so there is nothing to integrate.

Will it monitor our agents?

No. It does not record calls, log keystrokes or score agents with AI. Managers see mood only as an anonymous team picture.

Where is our data held?

On EU hosting. AI answers are generated by Anthropic’s Claude, which processes requests under standard UK and EU transfer safeguards.

Is there a free trial?

Not at the moment. Get in touch and I’ll show you around.

How long does setup take?

You don’t install anything. I set up your workspace, and your admin adds your agents and your knowledge.

Does it help with Consumer Duty?

Yes. Calls are logged with outcomes, vulnerability and disclosures, and managers get a control room and a CSV export. Small groups are hidden to protect individuals, and the report is written without AI.

Can we run more than one brand?

Yes, on Enterprise.

Can we change tier later?

Yes. Every change goes through me, so you talk to a person.

Playbook

Workspace

Your setup

On the calls

On the floor

Answers during the call

Coaching, not surveillance

Motivation built in

Consumer Duty evidence

Knowledge that keeps up

Be one of ten.

Your data, walled off

EU hosting

Sign-in

An audit trail nobody can edit

Erasure

Accessible to agents