Semantic density: old wine, or a new vintage?
Giving AI agents enough context to carry out enterprise tasks has become a priority for firms building them, but it’s expensive. It might not have to be.
Data management teams training AI agents by steeping models in context and metadata have encountered a problem: context is expensive. The more words used to teach an agent, the more words the agent uses to answer a prompt, resulting in higher token bills.
Tony McManus, global head of enterprise data at Bloomberg, says this has created a complex balancing act when training AI agents, where the prompts must include the highest percentage of relevant information, without making the agent return
Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.
To access these options, along with all other subscription benefits, please contact info@waterstechnology.com or view our subscription options here: https://subscriptions.waterstechnology.com/subscribe
You are currently unable to print this content. Please contact info@waterstechnology.com to find out more.
You are currently unable to copy this content. Please contact info@waterstechnology.com to find out more.
Copyright Infopro Digital Limited. All rights reserved.
As outlined in our terms and conditions, https://www.infopro-digital.com/terms-and-conditions/subscriptions/ (point 2.4), printing is limited to a single copy.
If you would like to purchase additional rights please email info@waterstechnology.com
Copyright Infopro Digital Limited. All rights reserved.
You may share this content using our article tools. As outlined in our terms and conditions, https://www.infopro-digital.com/terms-and-conditions/subscriptions/ (clause 2.4), an Authorised User may only make one copy of the materials for their own personal use. You must also comply with the restrictions in clause 2.5.
If you would like to purchase additional rights please email info@waterstechnology.com
More on Data Management
Firms get AI access to LSEG data via DataBP admin layer
LSEG has expanded its agreement with DataBP to enable customers’ AI-based tools to request and evaluate data from the exchange and maintain more comprehensive usage records.
Rimes builds agents for exception management
The vendor aims to improve its data foundations to allow for greater agent autonomy.
The great migration: Firms work to integrate ISO 20022
Banks and asset managers are working to adopt the new messaging standard, but it’s easier said than done.
PostSig, CJC team up to run data management for London broker
In the first such deployment between the vendors, CJC will act as an outsourced data administrator using PostSig’s technology to manage the firm’s data.
Buy-side firms warn semantic layers not a quick fix for AI integration
Semantic layers promise to teach AI what a firm’s data means. Keeping that meaning current is the part that may get overlooked, tech heads warn.
Mind the stack: the foundation markets run on
Exchange data is an underappreciated component of the AI era, writes Jeff Kimsey.
BlackRock Aladdin prioritizing semantic layer for agentic AI
The asset manager’s tech platform is spending “a lot of time thinking about” semantics before it rolls out AI for a variety of tasks, says Aladdin exec.
The impact of market data and alt data converging
The Waters Wrap: As firms combine traditional data sources with proprietary research and alt-data signals, Max has some questions about where the data market is headed.
