The day before August 2025

🔥 What's New ?

👉 Google Cloud's specialized agents are in preview

We told you about the Data Engineering Agent announced at Google Next 2025 in our May watch, and now Google has announced a preview launch for another agent dedicated to data science:

Data Engineering Agent ( in Preview) which automates complex data pipelines via natural language prompts. Users can now describe their requirements ("Create a pipeline to load a CSV file, clean up these columns, and join it to another table") and the agent generates and orchestrates the entire workflow in BigQuery.

Data Science Agent (also in Preview) which transforms the Colab Enterprise Notebook experience with complete autonomous analytical workflows. Powered by Gemini, it automatically triggers exploratory data analysis (EDA), cleansing, featurization, ML predictions, creating a plan, executing code, reasoning about results and presenting its conclusions.

👉 At Starfox Analytics, we support our customers in adopting Google Cloud's specialized AI agents to transform their data workflows, taking full advantage of the Data Engineering Agent's automation capabilities and the Data Science Agent's advanced analytics. Our expertise enables companies to optimize their data strategy by integrating these cutting-edge technologies to rapidly generate business value while reducing technical complexity.

ℹ️ For more information: New agents and AI foundations for data teams | Google Cloud ...

👉 Databricks integrates the new OpenAI open-source models (gpt-oss 20B and 120B)

Databricks has announced a major strategic partnership with OpenAI to launch two new open-weight models: gpt-oss 20B and gpt-oss 120B, available natively on the Databricks platform across AWS, Azure and GCP via the Foundation Model API.

These models represent a very important evolution in the enterprise AI ecosystem, offering advanced reasoning and tooling capabilities while maintaining the flexibility and customization options that proprietary models can't match. Companies can now deploy GPT-level models with full control over their data and deployments.

Native integration in Databricks Lakehouse enables organizations to leverage these models directly on their data without external transfer, guaranteeing security and governance.

ℹ️ Find out more: Introducing OpenAI's New Open Models on Databricks | Databri...

👉 At Starfox Analytics, we actively support our customers in building their data platforms by adopting effective data strategies to optimize their business.

👉 Qbeast raises $7.6M to revolutionize the efficiency of open data platforms

Qbeast, a data optimization platform, has secured $7.6 million in seed funding to develop its analytics solutions for open Lakehouse. Their multi-dimensional indexing technology optimizes the use of computational resources in open data formats such as Delta Lake, Iceberg and Hudi.

Tests in production environments show a 6-fold improvement in query performance, and a reduction in computing costs of up to 70%. This solution meets the challenges of today's data ecosystem, which is faced with rising cloud costs proportional to the growth in data volumes.

Qbeast technology is compatible with the main analysis platforms (Databricks, Snowflake, AWS) and integrates without modifying existing workflows, simplifying adoption and enabling rapid optimization of workloads.

ℹ️ Find out more: Qbeast

👉 dbt Labs forges strategic partnership with AWS

dbt Labs has announced the signing of a Strategic Collaboration Agreement (SCA) with (AWS). This agreement will extend technical integrations across various services, including Amazon Redshift, Amazon Athena and Amazon SageMaker Lakehouse, while expanding availability on AWS Marketplace to simplify access and procurement for enterprise customers.

ℹ️ Find out more: dbt Labs Signs Strategic Collaboration Agreement with AWS | ...

​💡Tip of the month

Our "Tip of the month" section shares practical tips used daily at Starfox Analytics. These tips cover various Web Analytics tools to optimize your work. Don't hesitate to try them out and share them with your colleagues.

👉 Tracking Checker: a free tool to assess your site's tracking health

A few months ago, Stape launched Tracking Checker, a tool that lets you assess the overall health of website tracking in just two clicks, without going into the technical details.

The tool analyzes several elements, such as the tracking tools used, cookies with their suppliers, categories and storage locations.

To obtain an analysis report, simply enter the URL of the site and your e-mail address.

We tested the tool on a media site, and here's what one section of the analysis report looks like 👇

ℹ️ For more information, visit Website Tracking Checker by Stape

‍📖 Sharing Is Caring

Every month, our "Sharing is Caring" column presents an in-depth article on a current topic in Web Analytics. Our experts use their know-how and online resources to explore these topics in detail.

👉 Article of the moment: Understanding the difference between an ETL and an ELT‍

This month, discover our business case on the server-side migration of our customer Back Market :

ℹ️ Back Market migrates to server-side for +14% tracked conversions

❤️ Best resources and articles of the moment

• Anthropic has set up a platform for learning all the concepts of artificial intelligence: Anthropic Courses‍
• How to Deploy an MCP Server to Google Cloud Run : Build and Deploy a Remote MCP Server to Google Cloud Run in ...‍
• A detailed guide to the concept of the "Semantic Layer" in data modeling by MotherDuck : Why Semantic Layers Matter - and How to Build One with DuckD
• ‍Usercentrics explains what a cookie popup is: Understanding the cookie popup

😜 Miscellaneous

👉 Fidji Simo: A Frenchwoman at the helm of OpenAI.

At the age of 39, a Frenchwoman reaches the highest echelons of OpenAI, the world's most influential artificial intelligence company. Fidji Simo's ascent has been remarkable. On Thursday May 8, the Sète native was appointed by OpenAI boss Sam Altman to the position of "CEO of applications at the company that created the ChatGPT conversational assistant. She will assume these new responsibilities in a few weeks' time, while a successor is found at the head of Instacart, the e-commerce start-up she previously managed.

ℹ️ Find out more: Fidji Simo at Open AI: the fabulous destiny of a Frenchwoman dev...

🤩 Inside Starfox

Starfox Analytics launches Data Circle: the alliance of data excellence at the service of digital transformation

We've always believed that the best talent isn't to be found in offices, but in free spirits who refuse to compromise. Today, we're reinventing data & analytics consulting with Data Circle - a revolutionary model that fuses the autonomy of independent experts with the strength of a leading agency infrastructure.

For our customers, it's simple: you get instant access to the best experts on the market, united by a shared vision of excellence. No junior trainees, no interchangeable resources - just exceptional talents, selected through co-optation and sponsorship, who share our exacting standards. Each Circle member contributes not only his or her expertise, but also his or her ability to innovate, challenge and transform.

Recently, we orchestrated in 72 hours the assembly of a dream team of 6 data experts for Voyage Privé - data governance, advanced analytics, server-side tracking, data engineering, Snowplow architecture and BI - transforming a strategic emergency into an opportunity to completely reinvent their data ecosystem. This surgical agility, impossible in a traditional model, is now our standard.

Circle members benefit from Starfox's full technological arsenal: proprietary GA4/BigQuery query libraries, proven methodological frameworks, AI-powered technology watch, collaborative technical wikis, business case templates, privileged access to SFXDays events, a Data Lab for collective experimentation, and a gamified recognition system. This is the first time that a collective of freelancers has benefited from such an infrastructure - that of an established agency, without the constraints of a traditional structure.

We're not just creating a network. We're building the future of consulting: an ecosystem where individual excellence amplifies collective power, where each expert remains free while belonging to something bigger. Data Circle is proof that independence and structure are not opposites - they're complementary. And together, they will transform the way companies mobilize data expertise.