KredX Collections
Empowering Finance Teams to Collect Smarter, Not Harder
%
Reduced Delays
%
Increased Efficiency
%
Faster Resolution
%
Happy clients
Role
Lead Product Designer
Team
Product, Engineering, Data Science, UX Research, QA
Time Spent
3 month (MVP in 1.5 months, scaled by 3rd Month )
Overview
Our customers’ finance teams were overwhelmed. Their accounts receivable (AR) processes were fragmented, slow, and overly manual. Chasing payments meant juggling spreadsheets, emails, phone calls, and unclear ownership. Cash flow forecasting was more guesswork than science.
Problems
No clear way to prioritize follow-ups, Manual & inconsistent reminders , Zero visibility into dispute status Unpredictable DSO (Days Sales Outstanding).
Solution
We designed an AI-powered collections tool that: Predicts payment likelihood and prioritizes follow-ups, Automates customized dunning workflows (email, SMS, escalation), Centralizes dispute tracking for internal teams and customers, Surfaces real-time dashboards for performance and cash insights.
Who it's for
Finance teams across large and mid-sized enterprises issuing high volumes of invoices.
Research & Discovery
Accounts receivable is often overlooked in digital transformation. But it's a critical lever for cash flow. We spent time with AR teams, watched them work, mapped workflows, and documented inefficiencies. From email chases to SAP exports, we saw how broken the current system was—and how ripe it was for intelligent automation.
Why This Solution?
Accounts receivable (AR) is often treated as a back-office cost center, but delays in collections directly impact cash flow and growth. We saw a clear opportunity to elevate AR into a data-driven, strategic advantage.
Domain Deep Dive
We studied AR operations across fintechs, ERPs, and enterprise finance stacks—particularly B2B workflows, multi-party invoice lifecycles, and dispute bottlenecks.
User Research
We shadowed 10 collections agents and conducted 24 interviews across 6 client companies. Key Insights:1. Agents lacked prioritization logic—often going by gut or recency2. Reminder emails were manual, repetitive, and easy to overlook 3. No audit trail or visibility for disputes—leading to missed SLAs
Design Goals
We zeroed down on following key goals for MVP:1. Replace chaos with clarity—give agents a daily playbook2. Reduce DSO ( Dispute Settlement Outstanding ) through smarter outreach, not brute force 3. Make AI explainable and adjustable4. Provide management oversight without micromanagement
Designing the Solution
Smart Worklist
"Help me see what matters first." We built dynamic invoice cards that ranked accounts based on AI-predicted payment likelihood. Each card gave agents context at a glance: days overdue, due amount, escalation risk, and even a reason why the AI flagged it. No more guesswork—just clear, confident action.
Visual Workflow Builder
"Let me design my own outreach strategy." We replaced hardcoded logic with a no-code, drag-and-drop canvas. Admins could now create automated dunning flows with delays, channel options (email, SMS, WhatsApp), filters, and retries. It was like building a journey map—only this one ran itself.
Dispute Management (Internal + Customer Portal)
"Where’s that stuck? And why?" We created a single form structure that worked both internally and externally. Customers could raise disputes from their portal. Internally, finance teams could triage, tag, and assign resolution steps. Everyone finally had the same view.
Executive Dashboards
"Give me numbers I can trust." For managers and CFOs, we designed dashboards that tracked DSO trends, aging buckets, collector productivity, and dispute volumes. With real-time metrics, they could finally plan proactively instead of reactively.
Impact
Agents went from chasing at random to targeting accounts with the highest impact and success probability.
This gave teams control over tone, timing, and escalation, while maintaining consistency. Outreach became proactive and automated, cutting manual work and standardizing customer experience.
Hi-Fidelity Design
At this stage, the goal was to move from validated concepts to functional, scalable UI—bridging the gap between product ambition and operational reality. We focused on building an interface that could support high-volume tasks, empower different user personas, and maintain clarity even in the most complex cases.
Impact of Design Implementation
Dramatic Reduction in DSO
A logistics client who handled thousands of overdue invoices saw their Days Sales Outstanding (DSO) drop from approximately 62 days to 34 days within three months of using KredX Collections. We tracked this improvement by comparing ERP payment dates before and after implementation and saw nearly a 45% reduction—freeing working capital and boosting operational cash flow.
Substantial Efficiency Improvements
With AI-driven prioritization and automated reminder workflows, collections agents reclaimed an estimated 70% of their previous workload. We tracked time-use through internal dashboards and manual time logs. AI automation replaced repetitive tasks and data entry, empowering teams to handle double their earlier invoice volume without increasing headcount—directly supporting the “70% increase in operational efficiency” claim in the KredX CMS messaging
Forecast Accuracy Through Real-Time Insights
C-level finance professionals reported improved forecast accuracy by up to 20% after using real-time analytics dashboards. These dashboards display key metrics like DSO trends, collector KPIs, and aging analysis—freeing decision-makers from Excel-based projections to data-informed planning
Future Roadmap
Channel Intelligence for Smarter Outreach
Our next step is building channel intelligence that recommends the optimal medium per customer—learning from response data to improve reply rates and reduce friction.
Reflections & Learning
Building Trust in AI-Powered Systems
We discovered that enterprises only adopt predictive automation if they understand it. Inline explainability and override features significantly boost engagement and user trust. This aligns with the broader need for transparency highlighted in KredX’s product ethos
System-Oriented UX Design
Every feature—from invoice cards to workflows—was influenced by ERP/CRM integrations, compliance requirements, and data provenance expectations. Enterprise UX is as much about policy alignment and system architecture as it is about interaction design.
Cross-Functional Feedback Fuels Refinement
Our most valuable refinements originated from early collaboration with finance SMEs, QA leads, and data scientists. For example, filter presets and side-panel dispute editing only emerged after observing real user workflows. Embedding this feedback into our iterative cycles made our prototype not just usable, but truly effective.
Let’s Connect
"This case study shares an overview of my design process and impact at a high level. Some details have been kept brief to respect the nature of the work. If you’d like to learn more about this project or discuss my approach in greater depth, feel free to reach out—I’d be glad to share more in conversation."