Join us for an informative morning focused on the key announcements, technologies, and insights emerging from the Google Cloud Summit 2026.

The event will explore developments across Google Cloud, Gemini, artificial intelligence, cloud infrastructure, data platforms, and developer tools, while demonstrating how these technologies can be applied within real-world software, data, and business environments.

Your Data Warehouse Can Talk. Should You Believe It?

The session will be presented by Nomthandazo Tshuma, an experienced Senior Data and AI Engineer who specialises in connecting complex data infrastructure with advanced artificial intelligence solutions.

As a Google Developer Expert and Google Cloud Innovator Champion, Nomthandazo brings extensive architectural expertise in designing resilient data and machine learning pipelines, scalable data ecosystems, and production-ready AI solutions.

Her professional experience spans the telecommunications, retail, and consumer sectors, where she has supported organisations in transforming raw data into meaningful business value through real-time analytics, predictive modeling, and modern cloud technologies. She is also an active community leader, mentor, and conference speaker who advocates for inclusive artificial intelligence, knowledge sharing, professional development, and open-source innovation.

In Your Data Warehouse Can Talk. Should You Believe It? Nomthandazo will examine the opportunities and governance challenges associated with natural language analytics.

Conversational analytics is rapidly becoming a practical reality. However, relying on an agent’s response without appropriate verification is not an effective data governance strategy.

During this practical workshop, attendees will learn how to build a trusted conversational analytics solution from end to end. The session will include:

• Querying BigQuery directly from the Antigravity IDE using the Google Cloud Data Agent Kit

• Creating and configuring BigQuery Data Agents to process complex SQL queries and forecasting requests in natural language

• Publishing data agents for organisational use through Looker Studio dashboards

• Implementing row-level security within multi-tenant data environments

• Establishing grounding rules to reduce hallucinated or inaccurate metrics

• Requiring agents to execute SQL and return verified results rather than relying on stored or assumed information

Drawing on her experience implementing governed conversational agents over multi-tenant human resources data in production environments, Nomthandazo will share practical lessons for developing secure, reliable, and trustworthy data agents.

Attendees will gain an understanding of how to:

• Meet users within the tools they already use, including integrated development environments, conversational interfaces, and executive dashboards supported by a shared underlying agent

• Treat agent instructions as formal governance artefacts that are version-controlled, reviewed, tested, and maintained

• Require agents to execute SQL and return verified metrics rather than responding from memory

• Protect sensitive information through appropriate row-level security

• Introduce conversational analytics as an additional layer within an existing data architecture rather than as a replacement for established systems

As this is an interactive, hands-on workshop, attendees are kindly requested to bring a fully charged laptop, have access to an active Gmail account, and be prepared to code along during the practical demonstrations and activities.

Please note that this is a paid event. A non-refundable registration fee of R60 per attendee is required to confirm attendance and secure a place at the event.

The fee contributes towards event logistics, catering, and technical resources required for the practical session. Attendees will benefit from expert-led content, hands-on learning, practical demonstrations, networking opportunities, and insights that can be applied to real-world projects.