What was the impact of the AI solution?

The article says the solution delivered measurable efficiency gains across global sales and trading operations. It reduced IT time and effort consumed by repetitive dashboard creation, freed traders from manual data wrangling, and democratized secure data access.

Source

Evidence: Low25/100

Jefferies builds an agentic trade assistant with Amazon Bedrock Knowledge Bases and AgentCore

Use case typeTrading platform modernizationUpdated 5 days ago

Jefferies built an embedded natural-language trade assistant for equities traders to query millions of rows of trade and execution data across multiple systems. The solution uses Amazon Bedrock Knowledge Bases, Amazon Bedrock Guardrails, Anthropic Claude, Amazon EKS, Strands Agents, and MCP tools to generate SQL, retrieve schema context, run queries, and render visualizations. The roadmap includes a global rollout, enhanced audit capabilities, and Amazon Bedrock AgentCore features.

Organization: Jefferies
Industry: Finance
Location: United States
Published: July 2026

Reported outcomes

Strategic outcomes

Better decisions & insightReduced manual data wrangling for tradersCost efficiencyReduced IT effort for dashboard creationBetter decisions & insightDemocratized secure access to trading dataCustomer experience & trustFaster onboarding and relationship management capacityInnovation & cultureShifted decisions toward real-time analytics

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Jefferies
Provider: AWS
Maturity: Unknown
Linked source: AWS Machine Learning Blog

No explicit deployment-stage evidence found.

Customer identity supportedSource describes one deploymentMaturity evidence evaluated

Primary read

Use case focus

Showing 3 of 3

1Trading platform modernization
2Real-time analytics
3Decision support

Traders needed real-time insights from vast trade and execution data without relying on subject matter experts or lengthy IT dashboard cycles.

Jefferies built an agentic AI trade assistant with a conversational interface, schema retrieval from Bedrock Knowledge Bases, LLM-generated SQL, MCP tools for multiple data sources, and security controls in Bedrock Guardrails.
The architecture uses an authentication/session layer on Amazon EKS and dedicated visualization engines rather than letting the LLM generate charts.

Technologies

Amazon Bedrock Amazon Bedrock Knowledge Bases Amazon Bedrock Guardrails Amazon EKS Amazon Bedrock AgentCore Anthropic Claude 3 Amazon Titan embeddings Strands Agents SDK Model Context Protocol

The article says the solution delivered measurable efficiency gains across global sales and trading operations.
It reduced IT time and effort consumed by repetitive dashboard creation, freed traders from manual data wrangling, and democratized secure data access.

Architecture

An embedded trade assistant widget in Jefferies' trading interface routes trader questions through an authentication/session service on Amazon EKS and a Strands-based query agent. The agent uses Amazon Bedrock Knowledge Bases with Titan Embeddings to retrieve schema and query context, then uses Anthropic Claude on Amazon Bedrock to generate SQL and orchestrate Model Context Protocol tools across in-memory grids, SQL sources, and FIX message files. Amazon Bedrock Guardrails adds PII filtering, content moderation, and policy alignment, while dedicated visualization engines render charts and graphs. Amazon Bedrock AgentCore is planned for future rollout.

Sources & evidence1

Evidence: Low25/100Evidence strength