Project snapshot
Client
AVER, Latvia-based real estate agency
Request
Automate calling and lead qualification operations and integrate with the client’s existing sales processes and CRM workflows
Solution
A custom AI calling agent built on Oxagile’s technology accelerator


The impact in numbers

The business case that brought AVER to us

“For a company our size, we didn’t think this level of customization and delivery speed was possible. The AI calling agent has been contributing to at least three solid deals a month.”
— Austra Kausiniece, CEO of AVER

Our goal is to increase the volume of calls we handle without adding more workload to our existing sales team. We need an AI calling agent that can manage both inbound and outbound calls, qualify leads, schedule follow-ups, and work seamlessly with our existing CRM and sales flows. Most importantly, the AI needs to understand our sales methodology and adapt to the scripts’ style our team generates.”
— Sales Team Manager, AVER
How we balanced speed, customization, and budget
We adapted AVER’s sales training guidelines into AI-ready scripts and conversation flows that reflected their tone of voice and communication strategy. This approach accelerated delivery, keeping the project within the planned development budget.
— Dmitry Karpovich, Founder of Oxagile

Meet the AI calling agent
Automated calling
The AI agent handles both incoming and outgoing calls.
Lead qualification
The agent conducts pre-scripted and personalized conversations, qualifies prospects, and identifies leads that require further sales attention.
Automated follow-ups
Scheduled follow-ups allow prospects to be contacted without requiring manual intervention from the sales team.
CRM integration
The AI calling workflow is connected with the client’s CRM for relevant leads and call information to flow into the existing sales process.
Connecting the AI calling agent to the client’s CRM
The AI calling agent integrates with AVER’s CRM to keep customer context and call outcomes in sync throughout the funnel. Here’s how we translated the sales challenges caused by manual processes into integration requirements:
- New leads pulled from the CRM or connected source
- Funnel stage used to select the calling flow
- Qualification captured during the call
- Outcome and transcript written back to the CRM
Turning sales expertise into AI-ready conversations
Analyze existing sales materials
Oxagile worked with the client’s team to turn their sales knowledge, messaging, qualification criteria, and techniques into a clear foundation for the AI agent.
Build AI-ready conversation flows
We transformed the educational materials into structured call scripts and conversation flows tailored to the agent’s role and the client’s sales methodology.
Test with real prospects
The team tested the agent within small groups of prospects, with call recordings manually reviewed to identify gaps and opportunities for improvement.
Refine and retest
Oxagile incorporated feedback from the client’s sales team, optimized the scripts and responses, and continuously retested the agent based on real interactions.
Optimizing real-world conversations
Challenge
Live customer calls revealed cases that were difficult to anticipate during initial script development. Customers interrupted the agent, spoke over it, responded unexpectedly, or called from noisy environments.
These situations could cause the agent to misinterpret the conversation, follow the wrong dialogue path, or respond less naturally than expected. Standardizing the agent’s behavior across such scenarios became an ongoing challenge.
How we solved it
Our team introduced a continuous call-review and optimization process. We manually reviewed call recordings and transcripts to identify specific failure points, such as interruptions, misunderstood responses, awkward transitions, and incorrect dialogue paths.
Then we combined these findings with feedback from the sales team and translated them into targeted updates to the agent’s prompts, scripts, and response logic. Finally, AVER tested new script versions in live calls.

Delivering real value and client satisfaction

“Oxagile quickly understood how our agency works and built an AI calling agent around our actual sales process. It now handles our first calls, qualifies leads, and gives our agents everything they need to follow up. The AI call center has been contributing to at least three solid deals a month.
What impressed us most was how fast Oxagile delivered such a customized solution. They worked with us as a true partner, continuously improving the system based on our team’s feedback. For a company our size, we didn’t think this level of customization and speed was possible.”
— Austra Kausiniece, CEO of AVER

Today, we are moving toward a much more outcome-driven model: focusing first on the business value we can create and then aligning our commercial model with them. That’s why we proposed an outcome-based pricing model, because we believe we should work in synergy with the value we deliver.”
— Dmitry Karpovich, Founder of Oxagile







