Π
PYTHYA
Contact Center Performance Intelligence
Incorrect username or password for this account.
+ Or explore a prospect sandbox (any industry & size)
🖧 Connect to your Pythya server
Connect to your server
Sign in and load your live account data.
← Back to sign in
Upload Your Own Data
Excel or CSV with one row per agent. Read entirely in your browser — nothing is uploaded to a server.
← Back to sign in
Prospect Sandbox
Build a full year of sample data for any kind of operation. You sign in as Country Manager.
← Back to sign in
CX
concentrix
Performance Intelligence
—
—
Live
Today
What needs you
Five minutes, in and out.
Loading…
Executive Performance Dashboard
Loading…
or to
—
Total Calls
—
Agents
—
Avg AHT
—
OSAT %
—
HERO %
—
Occupancy
📞 Calls and handling
🧑‍💼 Workforce & Schedule
✅ Quality & Customer Experience
Performance Leaderboard
Top 15 agents · voice, quality and survey score
AHT Distribution
Handle time per agent, against your own goal and floor. Green is at your goal, red is at your floor.
Transfer rate vs AHUF %
Top-left = risk · bottom-right = best
OSAT % vs HERO % — Customer Experience
Top-right = champions
Monthly Trend — volume, AHT & satisfaction
Change from month to month is shown below.
Performance Analytics · rule-based
What the Numbers Say
Analysing…
Best Performing Agents
Highest scores
Coaching Priority
Lowest scores — start here
Recommended Actions
This week's priorities
Performance Score Distribution
Agent count by band
Top 10 vs Bottom 10
Average KPI comparison
Agent 360
Performance radar
Six dimensions, scored 0–100
Where the handle time actually goes
Talk, hold and wrap for this agent against the team median.
All Metrics vs Team Average
Green above · Orange below
Compare Agents
vs
Every metric
▲ = winner
Radar
Scored 0–100
Four key metrics
Scaled to fit one axis
Scenario Planning · Single Agent
Performance Simulator
What coaching one agent could move, and what it costs.
Select Agent
What you would spend
Projections update as you move these
Coaching hours / month 0h
Conversations about work they already know how to do — not teaching a new skill.
Your own (TL) hours on this 0h
Which KPIs are you coaching?
Revenue per call (optional)
Leave blank and the cost is shown in calls and hours.
Now vs projected
Per KPI you selected
⚠ A range, not a promise.
Performance Score — Before vs After
KPI Change Projection
Coaching & events log
What you did, and anything that happened to the numbers.
Demo mode — nothing is saved. Connect to a live server to keep events.
Type
What kind
Date
Minutes
Agent
KPIs you worked on
What happened — in your own words
Write it however you like. It is never turned into a number.
Operations Planning · Team-Wide
Team Impact Simulator
What coaching part of the team could move, and what it costs.
What you would spend
Outcomes update as you move these
% of Team Coached 0%
Coaching hours / agent / month 0h
Conversations, not courses. See the panel below on why raising this does not move the projection.
Investment / Agent / Month ($)
Focus Areas
Projected Team Outcomes
Before vs projected averages
⚠ Team impact projections are scenario estimates.
Impact on Key Metrics
Automated Monitoring · Real-Time
Smart Alert Center
Scanning…
Alert Distribution
Agent count per category
Risk Map — AHT × Transfer Rate
Top-right = highest risk
Executive Briefing
Executive Story
Natural-language intelligence briefing — written for leadership
Performance Scorecard
Normalised radar — team averages
Export Reports
Decision Pipeline · AI-assisted, human-decided
What should I do about this agent?
Five steps, from reading the notes to recording what happened.
Agent
Spread your hours across the team
Set your limits. We show where the hours buy the most, agent by agent.
Coaching hours you can spend
Cash limit (optional)
Agents to consider
Demo mode. This screen needs a live server, because every step is recorded. Connect to a server to use it.
1 · Read the notes
Your model reads this agent's coaching log and recorded disruptions, and turns the free text into labels.
2 · Confirm what it read the guard
Nothing the model said is used until you accept it. Correct anything that is wrong.
3 · The options, costed and ranked
Every action the numbers can support, ranked.
Hours to consider
Your own hours
Hours you can spare
Let the model explain
4 · Your decision
You decide. If you change it or refuse it, say why.
5 · What actually happened
Come back after the next period closes and record what happened.
Is the review real?
How often you change or refuse what was recommended.
What the model does and does not do
What it writes, and what it never writes
It does: turn your team leaders' written notes into structured labels, and explain the chosen option in language a person will actually read.
It does not: produce a single number. Every figure here comes from your own data and cited research.
And it cannot: invent an option. It picks from the list above. If it names something that is not on the list, the platform discards it and tells you it did.
Turn the model off in step 3 and only the explanation disappears.
EVIDENCE
Where every number comes from
Each finding below, what it is used for, and what it does not cover.
QUALITY · AUDIT-LEVEL, COUNTED
Evaluations
Score real interactions against your own form. Every number here feeds the same scoring, simulator and briefs — with a counted n.
Pythya