AI case study

CandidlyConversation steering

23% of turns were disengaging — Cait only found out after each conversation ended. State inference now runs live and adjusts mid-turn.

Published

Key results

Prediction AUC
0.90
Engaged Turn Rate Increase
64%
Disengaging Turn Decrease
11%

Result highlights

Unlock 3 result highlights

The story

Context

An AI financial guidance platform helping employees navigate high-stakes decisions on debt payoff, savings, retirement, and benefits, delivered through employer benefit programs.

Challenge

Cait, the company's AI financial planner, could only be evaluated after each conversation ended, with no way to steer toward resolution while still...

Solution
Unlock full story

Quotes

Unlock 3 more quotes

The company

Student loan debt management and financial wellness platform.

IndustryFinancial Services
LocationNew York, NY, USA
Employees51-250
Founded2016

The vendor

Framework and developer platform for building LLM-powered applications.

IndustrySoftware & Platforms
LocationSan Francisco, CA, USA
Employees11-50
Founded2022

Use case

Candidly's Conversation steering is part of this use case:

Conversational AI
32 case studies(-12% YoY)
Proven impact?
LowModerateVery Strong
3.6Moderate
1.4Lowwithin Financial Services
2.1Lowwithin Product Engineering

Similar Case Studies

Related implementations across industries and use cases

35 AI case studies in Conversational AI

280 AI case studies in Financial Services

608 AI case studies in Product Engineering