How Send Payments gave 40 hours a week back to its team with AI agents
A connected AI workforce now handles customer ops, compliance, and CRM admin around the clock.
"If we get 10 leads in for the day, 10 responses get out and each one of them is unique."
Ryk Neethling
CTO, Send Payments
What changed
Before Relevance AI, Send Payments was stuck in a cycle of manual bottlenecks that slowed growth and burned out the team.
"We've absolutely moved from AI as a potential to an AI-first mindset in everything we do."
For Send Payments' leadership team, the breakthrough was not just automation. It was the ability to create agents that could understand context, make decisions, and be trusted across real operational workflows.
Sam — Customer Response
Leads outside Australian business hours waited 12+ hours for a first reply. The team couldn't cover every time zone, and generic autoresponders were losing deals.
Sam handles inbound communication in-market and in-time-zone, interpreting context from the CRM and generating a unique, personalised response — not a template.
24/7 coverage without the team being online. Every lead gets a context-aware response within minutes, regardless of time zone.
Barry — Call QA & Compliance
As a regulated financial services business, every customer conversation needs to be recorded and reviewed. QA was manual, fragmented across spreadsheets, and couldn't keep up with call volume.
Barry reviews every call for QA signals and compliance requirements. When a potential complaint is detected, it's automatically flagged and routed to Karen for escalation.
Continuous compliance coverage across 100% of conversations — with less manual review overhead and a stronger operational safety net.
CRM Automation Agent
Account managers spent roughly 8 minutes after every call typing notes into the CRM. Across 300+ conversations a week, that added up to ~40 hours of pure admin.
An agent that understands the CRM schema, listens to what changed in each conversation, and updates records automatically — capturing the right fields without manual input.
40 hours a week returned to the team. Account managers focus on selling and supporting customers instead of data entry.
Connected AI Workforce
Individual agents solved individual problems, but the real bottleneck was the gaps between workflows — handoffs, escalations, and follow-ups that still required human coordination.
The team linked agents together so outputs from one workflow trigger the next: Sam's response feeds Barry's QA review, Barry's flags route to Karen, and CRM updates flow automatically.
A scalable path from point automation to end-to-end process automation — with non-technical team members building their own agents within weeks.
What started as automation became an AI-first operating model
Send Payments began with targeted agent use cases. Once those workflows proved themselves, the team expanded into a broader AI workforce — with non-technical members building their own agents within weeks.
Customers received replies that felt immediate and human, even outside local business hours
Each workflow had an owner and a clear job, from customer response to QA to CRM admin
The business moved beyond point automation into full process automation
"It's way more customizable than just 'here is an agent'. We can build exactly what we need."
Ready to build your AI workforce?
See how Relevance helps teams move from manual ops to AI-first — not just single automations.