Kelvata Analytics Hub
Welcome to Kelvata Analytics Hub
Kelvata Analytics Hub is a cloud-based Debt Collection Intelligence Platform that gives financial institutions and collection teams AI-powered insights to recover more — faster. Upload your portfolio, let the platform score and segment it, and focus your collectors exactly where they'll have the most impact.
Pre-requisites
Kelvata Analytics Hub is fully cloud-based — no software installation needed. All you need is a modern browser and a stable internet connection.
| Requirement | Specification |
|---|---|
| Hardware | Any PC, laptop, or tablet with 4 GB RAM or more |
| Browser | Google Chrome, Microsoft Edge, or Safari (latest version) |
| Internet | Stable broadband connection — 1 Mbps or faster recommended |
| File Formats | .xlsx or .csv, maximum 50 MB per upload |
Request a Demo
Access to Kelvata Analytics Hub starts with a demo request. The Kelvata team reviews each application, provisions your organisation's account, and sends you login credentials once approved.
Step 1 — Visit the Demo Request Page
Go to kelvataanalyticshub.com and click the "Request a Demo" button on the homepage. This opens the demo application form.
Step 2 — Fill In Your Details
The demo request form at kelvataanalyticshub.com.
The demo request form asks for the following information:
| Field | Required? | Notes |
|---|---|---|
| First Name | Required | Your given name. |
| Last Name | Required | Your surname. |
| Work Email | Required | This is where your access credentials will be sent. |
| Phone / WhatsApp | Optional | A number the Kelvata team can use to follow up. |
| Company / Organisation | Required | Your company or institution's name. |
| Industry | Optional | Choose the closest match: Banking & Finance, Debt Collection / Credit, Insurance, Retail & FMCG, Logistics & Supply Chain, Healthcare, NGO / Non-profit, Government / Public Sector, or Other. |
| Team Size | Optional | 1 to 10, 11 to 50, 51 to 200, or 200+. |
| What are you looking to solve? | Optional | BI Dashboard Development, Predictive Analytics & Forecasting, Credit Risk & Collections Analytics, Data Cleaning & Transformation, Excel / Power BI Automation, Training & Consultancy, or "Not sure yet — show me everything". |
| Anything else you'd like us to know? | Optional | Free-text field for any extra context about your needs. |
Click Book My Free Demo to submit. A confirmation screen titled "You're Booked In!" appears, confirming the Kelvata team has received your request.
Step 3 — Awaiting Approval
The Kelvata team reviews every request manually, usually within 1 business day. During this time:
- Your organisation's account is provisioned on the Kelvata platform
- A dedicated subdomain or login URL is set up for your team
- Your initial Admin account is created
Step 4 — Receiving Your Access
Once approved, you'll receive an email with:
| What You Receive | Details |
|---|---|
| Your login URL | A unique link for your organisation, e.g. app.kelvataanalyticshub.com |
| Admin email | Your business email confirmed as the Admin username |
| Temporary password | A one-time password — you will be prompted to change it on first login |
| Getting started guide | A quick-start PDF and link to this documentation |
Partner Login
Once you have received your login credentials from the Kelvata team, you can access the platform at any time from any modern browser.
The Partner Sign In page.
Open your browser and navigate to your organisation's Kelvata URL (provided in your welcome email).
Type your business email address and password, then click Sign In.
On your very first login, you'll be prompted to set a new permanent password. It must be at least 12 characters and include an uppercase letter, a lowercase letter, a number, and a special character.
Partner Dashboard
After logging in, you land directly on the Portfolio Intelligence Dashboard — your command centre for every metric across your debt collection operation.
The Portfolio Intelligence Dashboard — sidebar, summary strip, model status banner and main area.
Layout at a Glance
| Area | What It Shows |
|---|---|
| Left Sidebar | Navigation grouped under Analytics, ML Models, and New Modules. Click any item to load that module. |
| Summary Strip (top) | Portfolio totals — total records, collectors, debt, amount collected, PTP rate, and recovery rate. |
| Model Status Banner | Shows which version of your AI model is active, how many files it was trained on, and its accuracy. |
| Main Area | The currently selected module's charts, tables, and KPIs. |
Dashboards & Insights
Every analytics module lives behind the sidebar of the Portfolio Intelligence Dashboard. Each module loads its own KPIs, charts, and tables — use the Days filter at the top of the dashboard and click ⟳ Refresh to recalculate any module for a different date range.
Portfolio Overview
The Portfolio Overview module as it appears in the Portfolio Intelligence Dashboard.
The landing module after login. It shows a row of KPI cards summarising your whole portfolio, a Daily Collections Trend line chart comparing amounts collected against outstanding debt over time, a Status Distribution doughnut chart, and a Top Clients table (Client, Accounts, Total Debt, Collected, Recovery %).
Below the charts, an Anomaly Alerts panel flags unusual activity in your portfolio. Click ⬇ Download Report to export the full anomaly list.
Promise-to-Pay Analysis
The Promise-to-Pay Analysis module as it appears in the Portfolio Intelligence Dashboard.
Tracks Promise-to-Pay (PTP) performance with KPI cards for PTP Rate, Total PTPs, PTP Amount, and Fulfilment. A PTP by Collector bar chart and a PTP Trend line chart show how promises and fulfilment change over time.
The table below breaks results down by collector — Collector, Calls, PTPs, PTP Rate, PTP Amount, Paid, and Fulfilment — so you can see which collectors are converting promises into actual payments.
Collector Performance
The Collector Performance module as it appears in the Portfolio Intelligence Dashboard.
Compares every collector's output side by side. KPI cards show Active collectors, the Top Performer, Avg PTP Rate, and Avg Collected. A Leaderboard bar chart ranks collectors by performance, and an Activity Volume chart shows their call and account volumes.
The table lists each collector with Accounts, PTP Rate, Collected, Avg Debt, and an Efficiency score — use it to identify who to coach or reward.
Portfolio Aging
The Portfolio Aging module as it appears in the Portfolio Intelligence Dashboard.
Shows how much of your portfolio is concentrated in older, harder-to-collect accounts. KPI cards display the percentage of the portfolio in the 0-30, 30-60, 60-90, and 90+ day buckets, plus the overall Average Age of outstanding debt.
Two bar charts — Count and Value — break the portfolio down across five age buckets: 0-30, 31-60, 61-90, 91-180, and 180+ days.
Customer Segmentation
The Customer Segmentation module as it appears in the Portfolio Intelligence Dashboard.
Groups your portfolio into behavioural segments so collection strategies can be tailored to each group. KPI cards show the number of Segments identified, the High Value and Contactable percentages, and the total Promises (PTPs) across the portfolio.
Each segment appears as its own card showing its name, account count, total debt, and share of the portfolio. A Segment Share doughnut chart visualises each segment's share of total debt, and an Avg Debt by Segment bar chart compares the average outstanding balance across segments.
Predictive Scoring
The Predictive Scoring module as it appears in the Portfolio Intelligence Dashboard.
Uses your AI model to score every account by risk and recovery likelihood. Filter the module by Customer Type — All, Conventional, Digital, or Business — using the buttons at the top, each showing the account count for that segment.
KPI cards summarise High Risk, Medium Risk, and Low Risk account counts, plus the model's Avg Score and PTP Accuracy. The High-Priority Accounts table lists the top 20 accounts ranked by Priority Index (0.1–0.9), with Risk Score, Loan ID, Contact Name, Type, Outstanding amount, recommended Strategy, Priority, Best Call Time, PTP %, and Pay %.
Use Download Scored Excel or Download CSV to export the full scored list for the selected customer type.
Roll-Rate Analysis
The Roll-Rate Analysis module as it appears in the Portfolio Intelligence Dashboard.
Shows how accounts migrate between delinquency stages over time. The Delinquency Bucket Distribution bar chart breaks the portfolio into six buckets — Current, 1-30 DPD, 31-60 DPD, 61-90 DPD, 91-180 DPD, and 180+ DPD.
A Self-Cure Accounts panel highlights accounts likely to resolve without intervention, showing the account count, total value, percentage of the portfolio, and a recommended action. An Early Warning Signals list flags emerging risk patterns, and the Roll-Forward Probability table shows, for each bucket, the percentage of accounts that Roll to the next bucket, Stay, or Self-Cure, along with the account count.
Recovery Forecast
The Recovery Forecast module as it appears in the Portfolio Intelligence Dashboard.
Projects how much you're likely to collect by the end of the month. KPI cards show MTD So Far, Forecast EOM, your Target and % of Target (if a target has been set), the current Daily Rate, and Days Left in the month.
An MTD vs Forecast line chart plots Actual, Forecast, and Target side by side, while a Monthly Trend bar chart shows collections over previous months. A Top Clients MTD table lists each client's collected amount with a progress bar showing their share of the month's total.
Best Time to Call
The Best Time to Call module as it appears in the Portfolio Intelligence Dashboard.
Recommends the call windows most likely to get an answer, based on historical call outcomes for the selected date range. The top three windows are highlighted as 🥇 Best Window, 🥈 2nd Best, and 🥉 3rd Best, each with its answer rate or call count.
The Performance by Time Band table breaks results into five bands — Early Morning (07–09), Mid Morning (09–11), Early Noon (11–13), Afternoon (14–16), and Evening (17–18) — showing Calls Made, Answered, Answer Rate, and a Recommendation for each. An Activity Heatmap visualises call volume by hour and day of week.
Payers Management
The Payers Management module as it appears in the Portfolio Intelligence Dashboard.
Identifies which accounts are actively paying and which are not. KPI cards show Payers, Non-Payers, the Payer Rate, and total Non-Payer Debt. Use the All Accounts / Payers Only / Non-Payers Only buttons to filter the table below.
Where data allows, a Monthly Payer Trend bar chart compares payer vs non-payer counts over time. A Declining Payers table flags debtors whose payments are trending down (with their monthly trend, last payment, and months tracked), while an Increasing Payers table shows the opposite. The Treatment by Disposition table recommends an action for each disposition based on account count and average balance. The main table lists every account with Debtor, Status, Outstanding, Collected, Age, Payer status, and a recommended Action.
Click Download Payer Report Excel (4 sheets) to export the full breakdown.
Contactability Analysis
The Contactability Analysis module as it appears in the Portfolio Intelligence Dashboard.
Measures how reachable your portfolio is by phone. KPI cards show Contactable and Non-Contactable account counts, the overall Contact Rate, and the Total number of accounts.
The Recovery by Time Band table shows, for each calling time band, the number of Accounts, amount Collected, Recovery %, and Contact %, so you can see which call windows actually convert to payments. The Flagged Non-Contactables table lists debtors who were called two or more times without an answer, showing their Attempts, the time bands already tried, a recommended alternative band, and their Outstanding balance.
Trigger Analysis
The Trigger Analysis module as it appears in the Portfolio Intelligence Dashboard.
Identifies what's driving spikes in daily collections. KPI cards show the Daily Avg amount collected, the Spike Threshold, and the number of Spikes detected in the selected date range. Use the Filter Disposition buttons to narrow the analysis to a specific call disposition.
The Daily Recovery Spikes line chart plots daily collections against the spike threshold (shown as a dashed line). The What Triggered Payments table breaks results down by disposition — Accounts, Collected, Avg/Account, and Recovery % — to show which call outcomes precede payment spikes. Action Recommendations cards then suggest a segment-specific action, each tagged with a priority (High/Medium/Low) and the number of accounts affected.
Click Download Trigger Analysis Excel to export the full breakdown.
New Customer Upload
The New Customer Upload module as it appears in the Portfolio Intelligence Dashboard.
Lets you enrich and score a brand-new customer book before it enters your main portfolio. The system estimates each account's age (from date of birth or Kenyan ID prefix), scores mobile number contactability by network, and — if your AI model has already been trained — applies its learnings to prioritise every account.
Reference tables show how scoring works: a Kenyan ID Number Patterns table maps ID prefixes to estimated age, risk tier, and behaviour profile (e.g. 6-digit IDs → 65+, Low risk, reliable payers; 8-digit IDs → under 25, High risk, highest default likelihood), and a Mobile Number Patterns table maps number prefixes to network and contactability (e.g. Safaricom 0700-0704/0710-0722 → 85% contactable, easy M-Pesa collection; Telkom 077x → 35%, very hard to reach).
Under Upload Files, choose a Collections File (required, .xlsx or .csv) and optionally a CRM File and Dialer File — these are merged on a common ID or mobile number column. Click Analyse & Score New Book → to process.
Results show KPI cards for Total Accounts, whether the book was AI Scored, and the number of Files Merged; a Mobile Network Breakdown table (Network, Accounts, Share, Avg Contact %, Ease); an Age Distribution chart; ID Risk Tier Distribution cards (Low/Medium/High/Unknown); and an Account Preview table of the first 50 scored accounts. Click Download Scored New Customer Book (Excel) to export the enriched, scored file.
Managing Users
Admins manage the whole team from the Team page — inviting new users, assigning roles, controlling module access, and suspending or removing accounts.
The Team page where admins invite, assign roles to, and manage users.
Invite a User
Click Team in the sidebar. The page lists all current users with their role, last login, status, and (for Admins) module access count.
A form opens with three fields: Full name, Email address, and Role.
Select Analyst, Team Leader, Admin, or Viewer from the Role dropdown (see Roles & Permissions below). The role controls which modules the user can access by default.
Click Send Invite. The new user receives a welcome email with their temporary password and a link to set up their account.
Roles & Permissions
| Role | What They Can Do |
|---|---|
| Admin | Full access — manage users, billing, all 13 modules, upload portfolios, and download reports. Typically the organisation's primary account manager. |
| Team Leader | Access to all analytics and ML modules. Can download reports. Cannot manage users or billing. |
| Analyst | Read-only access to modules assigned by Admin. Can download reports. Cannot upload files or manage the team. |
| Viewer | View-only access to specific modules assigned by Admin. Cannot download, upload, or manage any settings. |
Suspending or Removing a User
Each row on the Team page has action buttons on the right. Click Suspend to immediately revoke a user's access without deleting their account — the button changes to Activate, which restores access at any time. Suspended users appear dimmed in the list.
Click Delete to permanently remove a user and their account. This cannot be undone.
Two-Factor Authentication
All users can add two-factor authentication (2FA) for extra security. We strongly recommend Admins enable this first — their accounts control billing and team access.
From your organisation's Kelvata URL, append /setup-2fa to the address and press Enter.
Open Google Authenticator or Authy on your phone and scan the QR code shown on screen.
Type the 6-digit code shown in your authenticator app into the verification field and click Confirm & Enable 2FA.
From this point, every login requires your password plus a fresh 6-digit code from the app.
Billing & Subscription
Admins can view the current plan, check usage limits, and upgrade via M-Pesa — all from the Billing page in the sidebar.
The Billing page shows your current plan, how many team members you can add, and your monthly upload allowance. Click a plan card to start an upgrade.
The Billing page showing the current plan and upgrade options.
| Plan | Price | Users | Uploads / Month | Included |
|---|---|---|---|---|
| Starter | KES 15,000 / mo | 3 | 5 | Full analytics + ML scoring, Priority support |
| Growth | KES 35,000 / mo | 10 | 20 | Full analytics + ML scoring, Priority support |
| Enterprise | KES 80,000 / mo | Unlimited | Unlimited | Full analytics + ML scoring, Priority support |
Every plan includes all 13 dashboard modules, AI/ML scoring, and priority support — the plans differ only in how many team members you can add and how many files you can upload per month.
Click the plan card you want — Starter, Growth, or Enterprise.
Type your M-Pesa phone number and click "Pay Now". An STK push prompt appears on your phone.
Enter your M-Pesa PIN to confirm. Your subscription updates automatically within 1–2 minutes.
Troubleshooting
Common issues and how to resolve them quickly.
| Problem | What to Do |
|---|---|
| File upload fails | Check the file is .xlsx or .csv and under 50 MB. Try downloading the sample file from the Upload page to confirm your connection works. |
| Dashboard shows no data | Make sure a portfolio has been uploaded and finished processing — you'll receive an email once scoring is complete. |
| Charts are blank | Try changing the Days filter and clicking ⟳ Refresh. Some modules need at least 30 days of data to show meaningful results. |
| Forgot password | Click "Forgot Password" on the login page. Check your spam folder if the email doesn't arrive within 5 minutes. |
| Scoring seems wrong | Your AI models retrain automatically each time you upload new data. Check the model status banner to see the current version and accuracy. The more data you upload over time, the sharper your predictions become. |
| Cannot see a module | Your administrator controls which of the 13 modules each team member can access. Ask them to check your access from the Team page. |
| M-Pesa payment not confirming | Wait 5 minutes and refresh the Billing page. If still pending, contact support@kelvataanalyticshub.com or WhatsApp 0701 418 508 with your M-Pesa transaction code. |
| Account locked | After 5 failed login attempts, accounts lock for 15 minutes. Wait and try again, or ask your administrator to unlock it immediately. |
| Demo request not approved | Check spam folder for the approval email. If no email after 2 business days, contact the team directly via email or WhatsApp. |
Support Details
Need help? The Kelvata team is available via multiple channels.
| Channel | Contact |
|---|---|
| support@kelvataanalyticshub.com | |
| 0701 418 508 | |
| 🌐 Website | kelvataanalyticshub.com |
Visit kelvataanalyticshub.com to learn more.