Customer support · Production

LSEG uses Amazon Q Business to enhance LCH post-trade client services query handling

The problem

LCH's client services team had to manually refer to detailed service and policy documentation to answer complex member queries across diverse asset classes. The team relied on product FAQs and an in-house knowledge center, but needed faster and more accurate information retrieval as their customer base grew.

Workflow diagram · grounded in source
1
CS agent submits member query
trigger
“a web-based interface that serves as a gateway for their internal client services team to interact with the Amazon Q Business API”
2
RAG retrieval and answer generation
ai_action
“Amazon Q Business saved time in generating answers, including summarizing documents, retrieving answers to complex Member enquiries, and combining information from different data sources (while providing in-text citations to the data sou…”
3
Answer returned to CS agent
output
“The Amazon Q Business application returned answers within a few seconds for each question”
4
Claude v2 scores responses against golden answers
validation
“The Amazon Bedrock service uses Anthropic's Claude v2 model to validate the Amazon Q Business application queries and responses against the golden answers stored in the S3 bucket”
5
Admin reviews accuracy scores
human_review
“Anthropic's Claude v2 model returns a score for each question and answer, in addition to a total score, which is then provided to the application admin for review”
Reported outcome

Amazon Q Business provided LCH's client services team with quick and accurate answers to complex member queries, with responses delivered within a few seconds and high factual accuracy demonstrated in testing, saving time for live agents.

Reported metrics
Response timewithin a few seconds
Factual accuracy levelhigh level of factual accuracy
Agent time per questionsaves time for each live agent on each question
Reported stack
Amazon Q BusinessAmazon BedrockAnthropic's Claude v2Amazon ECSRAG
Source
https://aws.amazon.com/blogs/machine-learning/london-stock-exchange-group-uses-amazon-q-business-to-enhance-post-trade-client-services?tag=soumet-20
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Amazon Q Business provided LCH's client services team with quick and accurate answers to complex member queries, with responses delivered within a few seconds and high factual accuracy demonstrated in testing, saving…

What tools did this team use?

Amazon Q Business, Amazon Bedrock, Anthropic's Claude v2, Amazon ECS, RAG.

What results were reported?

Response time: within a few seconds; Factual accuracy level: high level of factual accuracy; Agent time per question: saves time for each live agent on each question (source-reported, not independently verified).

How is this customer support AI workflow structured?

CS agent submits member query → RAG retrieval and answer generation → Answer returned to CS agent → Claude v2 scores responses against golden answers → Admin reviews accuracy scores.

Related customer support cases