PRONTO XI CRM
An AI-Powered, Role-Based CRM That Grew the Business
Success Matrix
35% increase in module upgrades (licensing) . AI capabilities embedded directly into daily workflows
Why
Pronto Xi's CRM was a records system — it stored what users typed in, but did nothing with it. Sales reps logged activity into a void; managers chased updates manually; the platform knew everything and surfaced nothing. Meanwhile the market moved to AI-native CRM, and Pronto risked looking a generation behind.
How
I rebuilt the CRM around a single idea — the system should do the thinking the user shouldn't have to. Role-based dashboards, AI surfaced directly in context, and a clean information architecture that replaced buried menus with what each role actually needs first.
What I solved
Turned a passive records system into an active decision-support platform — lead scoring, next-best-action, forecasting, and automated activity capture, all delivered through role-specific workspaces. The redesign drove a 35% lift in module upgrades and opened a new small-to-medium business market Pronto hadn't previously won.
THE PROBLEM
The old CRM treated every user identically and every record equally. A sales rep, a sales manager, and a service agent saw the same dense interface, the same undifferentiated data, the same manual workflows.
Three failures compounded:It captured, but didn't compute. Reps entered data the system never turned into insight — no scoring, no prioritisation, no forecasting.It served no role well. One generic layout meant everyone navigated past what they didn't need to reach what they did.It was falling behind the market. AI-native competitors were setting a new baseline; Pronto's CRM felt static beside them.
COMPETITOR ANALYSIS
I benchmarked against the platforms Pronto's prospects were actually evaluating:
1.Salesforce — the enterprise benchmark, with Agentforce driving lead scoring and autonomous agents. Powerful, but complex and expensive to implement.
2.Microsoft Dynamics 365 — deep Microsoft-ecosystem integration and Copilot for Sales; strong in large enterprises already on the Microsoft stack.
3.HubSpot — the accessibility play: a unified, fast-to-adopt interface winning mid-market and growth-stage teams.
4.Zoho, Freshworks, Pipedrive — aggressive on price and simplicity in the SMB segment.
The strategic gap I designed into: the incumbents win large enterprises but are heavy, costly, and slow to implement. Pronto Xi already had ERP depth — CRM plus operations on one platform, the same architecture Odoo and NetSuite compete on.
My job was to make the CRM feel as modern and AI-capable as HubSpot, while leaning on Pronto's ERP-integrated advantage.
That combination is what opened the small-to-medium business market — modern AI experience, without Salesforce-scale cost and complexity.
RESEARCH
A research-dense discovery phase, because a multi-role platform can't be designed from assumption:
-
Stakeholder and user interviews across sales, service, and management
-
Contextual inquiry — observing how reps and managers actually worked, not how the system assumed they did
-
Competitor benchmarking (above)
-
Heuristic evaluation of the existing CRM
-
30 usability and user testing sessions across roles, spanning wireframes through to high-fidelity prototypes
The insight that reframed everything: users weren't struggling because data was missing — they were struggling because the system made them do all the thinking themselves. The fix wasn't more fields. It was intelligence.
MULTI-ROLE JOURNEY MAP
The redesign lives or dies on serving distinct roles from one platform. I mapped three:
One CRM. One design system. Three genuinely different jobs — each served by a workspace configured for it, not a generic layout everyone fights.
SALES REP — "What should I do next?" Lands on a prioritised pipeline, AI lead scores surfacing the deals worth attention first → activity auto-captured, no manual logging → next-best-action recommended in context → acts, instead of administrating.
SALES MANAGER — "How is the team tracking?" Lands on team performance and AI forecasting → spots an underperforming rep or at-risk deal via surfaced signals, not manual digging → coaches based on evidence → forecasts with confidence, not guesswork.
SERVICE AGENT — "What needs me first?" Lands on a prioritised case queue → full customer context assembled automatically → resolves faster with the history already in front of them → escalates cleanly when needed.
THE AI LAYER
AI wasn't a bolted-on chatbot. It was embedded into the workflows users were already in:
​
-
Lead scoring — surfaces which deals deserve attention first, so reps stop guessing.
-
Activity capture — logs interactions automatically, ending manual data entry.
-
Next-best-action — recommends the specific next step, in context, at the point of decision.
-
Forecasting — gives managers evidence-based pipeline predictions, not gut feel.
-
Ask Pronto AI — a conversational layer to query the CRM in plain language.
-
Role-based dashboard intelligence — the same AI engine surfaces different insights depending on who's logged in, so a rep, manager, and agent each see what matters to them.
​
The design principle throughout: the interface has to know who's looking and what they're responsible for before AI can be useful. Role-based architecture and AI weren't two features — they were the same decision.
DESIGN SYSTEM & IA
Scalable by design — built on the Pronto Xi hybrid design system (IBM Carbon principles + Pronto-native components), so every CRM screen inherited consistency and every new module could reuse the patterns.
Clean information architecture — replaced buried, undifferentiated menus with role-first navigation: what you need, where you'd expect it, nothing you don't.Reduced forms — inline actions and auto-capture replaced navigate-to-form workflows, cutting steps out of high-frequency tasks.
OUTCOMES
Business
-
35% increase in module upgrades — the redesigned CRM directly drove existing clients to upgrade.
-
New market opened — the modern, AI-capable, ERP-integrated experience won small-to-medium business clients Pronto hadn't previously competed for.
-
Leadership recognition — the CRM redesign was recognised internally as a driver of licensing growth and competitive positioning.
Experience
-
Faster task completion across all three roles in testing.
-
Manual data entry sharply reduced through automated activity capture.
-
Managers forecasting on evidence instead of guesswork.
PlatformRole-based, AI-ready CRM architecture reusable across the wider Pronto Xi platform.
CONCLUSION
This wasn't a redesign that made the CRM look better. It was a redesign that made the business grow. A 35% lift in module upgrades, a new small-to-medium business market opened, and a role-based AI system that gives every user — rep, manager, agent — exactly what they need to act, not just what the database happened to store.