Following on the September 2025 AI Product Series data, which examined tools used for in-house AI application development, ETR is releasing its second iteration of the AI Product Stack Report to better understand which tools are being used in combination with one another to support AI use cases.

 

This Product Stack Report analyzes the likelihood that two products would be used together in a given tech stack. It shows simply "If a respondent is using product A, they are XX% more/less likely to also be using product B." The analysis helps explain not just which AI products and features are being evaluated and used in organizations, but also which clusters of products are more often used together by companies to achieve their AI goals. 

 

• Snowflake is emerging as its own ecosystem. Outside of the three major public cloud vendors (AWS, GCP, Microsoft Azure), some of the strongest co-occurrences across product area combinations are between Snowflake products or between Snowflake products and other vendors. Some particularly strong co-occurrences can be seen between Snowflake Cortex and Snowflake ML users and Datadog, for instance.
• Respondents are most likely to use product offerings within a single public cloud provider's ecosystem or from those providers' closest partnerships. For example, those using Microsoft products in one area of AI in-house development are more likely to use Microsoft or OpenAI products in other areas. Google's ecosystem is especially tight-knit with high co-occurrences between its products.
• Growing co-occurrences involving Databricks, Snowflake, Prometheus, and major cloud vendor products. Since the March 2025 survey, some co-occurrences have seen notable growth, suggesting organizations are increasingly using certain tools in combination. Nearly 50-percentage-point gains in co-occurrence can be seen between Microsoft Azure Machine Learning and Databricks DBRX and Snowflake Arctic. Prometheus saw sizable growth in co-occurrence between other observability products like Elastic as well as with Anthropic Claude users.
• Often, respondents use multiple products in the same product area (e.g., they have multiple language/foundation model providers). Common combinations within single product areas include Anthropic Claude + Meta Llama; Elasticsearch + Redis; and Dynatrace + Splunk; among others.

Looking for more AI Insights? Fill out the form below to be emailed a summary of our latest AI Product Series findings.