AI Engineer
← All organizations

Financial information, technology and media

Bloomberg

Bloomberg provides financial data, news and software for investment professionals. The Bloomberg Terminal combines market information, trade execution and messaging, while its enterprise products support investment analysis. PORT Enterprise connects permissioned private-fund data through a Canoe integration for analysis across asset classes. ASKB adds a conversational AI interface to the Terminal, extending how professionals interact with its information and tools.

Michael Bloomberg founded the company in 1981 with Thomas Secunda, Duncan MacMillan and Charles Zegar; Vlad Kliatchko is its CEO. Its financial-language research includes BloombergGPT, a 50-billion-parameter model introduced in 2023. The research combined 363 billion tokens from Bloomberg’s financial sources with 345 billion general-purpose tokens, exploring how specialized training could improve financial tasks while preserving general language capabilities.

Bloomberg’s terminals served more than 300,000 finance professionals globally in 2026. Separately, Bloomberg Media reported over 707,000 paying subscribers in February 2026. Bloomberg also signed an agreement in July 2026 to acquire Canoe Intelligence, which automates private-markets data collection and delivery. The transaction remained proposed and under initial ACCC assessment in August 2026.

www.bloomberg.com ↗

2 talks

Newest first

2 speakers at AIE

Affiliations reflect their AIE appearances, not necessarily current employment.

Anju KambadurLei ZhangBloomberg’s head of technology infrastructure

Start here

  1. Challenges to Scaling Agents for Generative AI Products

    Start with Anju Kambadur’s talk for the financial-data context behind analyst-facing agents, including structured and unstructured sources and the evolution toward hybrid indexing.

    Anju KambadurAI Engineer Summit 2025

  2. What We Learned Deploying AI within Bloomberg’s Engineering Organization

    Read Lei Zhang’s talk for how MCP connects agents to operational telemetry and service topology, and where training and engineering leadership enter AI adoption.

    Lei ZhangAI Engineer Code 2025

Messages from the stage

Safeguards for composed agents

Kambadur explains how stochastic errors compound across composed LLM agents, motivating downstream safety checks, MLOps remediation, and circuit breakers.

Engineering work beyond code generation

Zhang discusses agents for maintenance, migration, refactoring, and incident response. Deterministic verification and growing pull-request and merge queues frame the challenges of turning generated code into delivered changes.

Affiliations reflect each recorded session, not necessarily current employment.

Company sources · checked 2026-08-28