Customer support · Production

Rocket Companies streamlines home buying with Amazon Bedrock Agents

The problem

Home buying remains overwhelming for many clients, and Rocket needed to deliver the same trusted, expert guidance clients expect from its mortgage bankers at any hour and on any digital channel.

Workflow diagram · grounded in source

Client initiates chat

trigger

“Client initiation: The client uses the chat function within Rocket's mobile app or web page”

Request routed to domain agent

routing

“Agent routing: The AI Agent API routes the request to the correct Amazon Bedrock agent based on static criteria, such as web or mobile property that the client entered the chat through, or the use of LLM-based intent identification”

Agent task decomposition

ai_action

“Agent processing: The agent breaks the task into subtasks, determines the right sequence, and executes actions and knowledge as it works”

Knowledge base retrieval and action

integration

“Task execution: The agent uses Rocket data in knowledge bases to find info, send results to the client, and perform actions to get work done”

Guardrails enforcement

validation

“Guardrails: Enforce Rocket's responsible AI policies by blocking topics and language that deviate from the goals of the experience”

Escalation to human banker

human_review

“Rocket implemented uncertainty thresholds using confidence scores and specific keyword triggers to detect when an interaction might require human assistance. In those cases, Rocket AI Agent proactively transitions the session to a live s…”

Reported outcome

Since launching Rocket AI Agent, clients who interact with it are three times more likely to close a loan.
Rocket saw an 85% decrease in transfers to customer care and a 45% decrease in transfers to servicing specialists, while 68% of clients gave high satisfaction ratings.

Reported metrics

loan close rate vs non-AI-agent usersthree times more likely to close a loan

Conversion rate from web traffic to closed loansThreefold increase

Transfer to customer care85%

Transfer to servicing specialists45%

Show all 5 reported metrics

loan close rate vs non-AI-agent usersthree times more likely to close a loan

conversion rate from web traffic to closed loansThreefold increase

transfer to customer care85%

transfer to servicing specialists45%

client high satisfaction ratings (CSAT)68%

Reported stack

Amazon Bedrock Agents

Source

https://aws.amazon.com/blogs/machine-learning/how-rocket-streamlines-the-home-buying-experience-with-amazon-bedrock-agents?tag=soumet-20

Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Since launching Rocket AI Agent, clients who interact with it are three times more likely to close a loan.

What tools did this team use?

Amazon Bedrock Agents.

What results were reported?

loan close rate vs non-AI-agent users: three times more likely to close a loan; Conversion rate from web traffic to closed loans: Threefold increase; Transfer to customer care: 85%; Transfer to servicing specialists: 45% (source-reported, not independently verified).

How is this customer support AI workflow structured?

Client initiates chat → Request routed to domain agent → Agent task decomposition → Knowledge base retrieval and action → Guardrails enforcement → Escalation to human banker.

Since launching Rocket AI Agent, clients who interact with it are three times more likely to close a loan.Rocket saw an 85% decrease in transfers to customer care and a 45% decrease in transfers to servicing specialists, while 68% of clients gave high satisfaction ratings.