Generative AI fields now available in ECS allowing parity and compatibility with OTel

OTel and ECS help users simplify and normalize GenAI monitoring across different GenAI vendors and data sources.

blog-ecs-otel.jpg

At Elastic, we developed the Elastic Common Schema (ECS) to make it easy to normalize data from different sources. Our documentation defines ECS as “an open source specification, developed with support from the Elastic user community. ECS defines a common set of fields to be used when storing event data in Elasticsearch, such as logs and metrics.” 

In April 2023, Elastic contributed ECS to the OpenTelemetry Semantic Conventions (OTel) to support the development of a cross-industry open schema. As an ongoing commitment, Elastic actively continues to contribute to OTel. We also merge important updates to OTel back to ECS since many Elastic users already have ECS workflows set up and may wish to run OTel and ECS in tandem.

With the rising use of generative AI (GenAI) and agentic AI, there is a growing need for fields that capture the context of GenAI models’ requests and responses.

Use cases include:

  • Observability for GenAI applications

  • Securing and protecting generative AI applications

  • Normalizing the logs from various generative AI vendors

  • Users benefit from OTel support, users benefit from turnkey log integrations that are compatible with ECS and OTel

GenAI fields in OTel and ECS

As introduced above, there are many use cases where users may wish to monitor their GenAI applications. To meet this demand, gen_ai attributes were added to OTel. This could include information on messages in a conversation, the setup for the large language models (LLMs), timestamps, token counts, and more.

Users can also normalize the logs from various sources on moderation. For example, they can use the GenAI fields in conjunction with security_rule (OTel) and rule (ECS) to capture any moderation activities in a GenAI application from guardrails that block keywords or potential malicious activity. 
See AWS Bedrock guardrails, Azure OpenAI filters, or the documentation for each vendor for details.

Generative AI fields in ECS are available in 9.1.0 in beta

Bringing OTel fields to ECS ensures compatibility with existing ECS workflows. At the same time, it ensures that users are able to benefit from the turnkey and neutral nature of a standardized definition for monitoring and securing their GenAI applications and infrastructure. To sync with the OTel GenAI fields, which are in the development phase, an initial batch of ECS GenAI fields are available in beta and subject to change.

You can start using these fields today in your GenAI workflows. Please, visit the ECS gen_ai fields page and the OTel gen_ai fields page.

The release and timing of any features or functionality described in this post remain at Elastic's sole discretion. Any features or functionality not currently available may not be delivered on time or at all.

In this blog post, we may have used or referred to third party generative AI tools, which are owned and operated by their respective owners. Elastic does not have any control over the third party tools and we have no responsibility or liability for their content, operation or use, nor for any loss or damage that may arise from your use of such tools. Please exercise caution when using AI tools with personal, sensitive or confidential information. Any data you submit may be used for AI training or other purposes. There is no guarantee that information you provide will be kept secure or confidential. You should familiarize yourself with the privacy practices and terms of use of any generative AI tools prior to use. 

Elastic, Elasticsearch, and associated marks are trademarks, logos, or registered trademarks of Elasticsearch N.V. in the United States and other countries. All other company and product names are trademarks, logos or registered trademarks of their respective owners.