Beyond the Prompt: Tuning Fabric Data Agents for Flawless Results
Data Engineers in Toronto August 2026 Semimonthly Meeting Topic: Beyond the Prompt - Tuning Fabric Data Agents for Flawless Results Abstract: As organizations race to deploy AI Agents, it quickly becomes clear that standard prompt engineering isn't enough. To achieve production-grade accuracy, you have to go beyond the prompt and fundamentally change how you feed your data layers to the AI. In Microsoft Fabric, your choice of data artifact directly dictates your agent's reasoning style and native language. In this session, we will deep-dive into how Fabric Data Agents interact with four distinct data paradigms: Semantic Models, Ontology, Data Warehouses, and Lakehouses. Moving past high-level theory, we will practically map the three pillars of data agent optimization across the data paradigms: Agent Instructions (System Prompts) Data Source Descriptions (Grounding) Example Queries (Few-Shot Prompting) You will walk away with an actionable blueprint showing you exactly how to write context rules, embed semantic metadata, and supply technical query examples to ensure your Fabric Data Agents deliver flawless, trusted results every time. Speaker: Thimantha Vidanagamage, Data Engineer at OMERS, Canada Speaker Profile: Thimantha has over 8 years of experience in Data and AI. He currently works as a Senior Data Engineer at OMERS in Canada, where he plays an active role in the organization’s adoption of Microsoft Fabric. He holds multiple Microsoft certifications, including Microsoft Certified: AI Transformation Leader and Fabric Data/Analytics Engineer Associate. Thimantha is also a co-organizer of the Toronto Data Professionals Community. The meeting is over Microsoft Teams, and the joining link is https://vip.dataengineersintoronto.org/webinar See you at the meeting!Making AI Work: Aligning Data, Teams, and Business Goals
AI adoption often stalls not due to a lack of tools, but because of what I call “frozen yoghurt syndrome”—an overwhelming number of choices that leads to confusion, misalignment, and stalled progress. This session offers a pragmatic framework for building a scalable data strategy that cuts through the noise and focuses on delivering business value. We’ll explore four key pillars: - Identify the Problem and Bigger Picture – Start with the business need. AI isn’t always the right answer; sometimes automation or simpler analytics are more effective. - Design the Strategy Around Data, AI, and the Use Case Together – Align initiatives with clear business goals and define success metrics that matter to both technical and non-technical teams. - Build the Foundation – Encourage cross-functional collaboration, select scalable tools, and document everything from data sources to model assumptions. - Create a Positive Feedback Loop Between Tech and Non-Tech Teams – Foster a shared product language, enable self-service access to data, and build trust through transparency and iteration. Attendees will leave with practical frameworks to ensure AI efforts are not just technically sound — but strategically aligned, collaborative, and sustainable.Data Mesh as the Foundation for AI/ML in Financial Services
Data Engineers in Toronto June 2026 Semimonthly Meeting Topic: Data Mesh as the Foundation for AI/ML in Financial Services Abstract: Financial institutions want AI/ML at scale, but brittle data pipelines, silos, and compliance demands slow progress. This talk shows how a Data Mesh—domain-oriented ownership, data-as-product, self-serve platforms, and federated governance—becomes the foundation for reliable, reusable ML features and trustworthy models. We’ll map mesh principles to FS use cases—fraud detection, risk, personalization—and show patterns for feature stores, lineage, quality, and access controls that satisfy regulators while accelerating delivery. Attendees will get a pragmatic blueprint: where to start, how to sequence capabilities, metrics that prove value, and pitfalls to avoid on the road from pilots to production. Speaker: Santosh Durgam, Data Engineering & Analytics Leader Speaker Profile: Santosh Durgam is a data engineering & analytics leader with 20+ years building governed, high-scale data platforms across retirement/401(k), broader financial services, and healthcare. He leads cross-functional teams that deliver production-grade data lakes, lineage-aware pipelines, and ML-enabled analytics on cloud—translating governance into measurable business outcomes. Recent speaking includes SQL Saturday Minnesota 2025, where he presented “From Ingestion to Insights: Building Robust Data Pipelines in AWS” to an in-person community audience. He has also contributed to international research forums and science conferences, and is invited to speak at ICDPN-2025 (International Conference on Data Processing & Networking), engaging practitioners and scholars on data engineering, governance, and analytics at scale. Santosh actively publishes/curates work via Google Scholar and shares practical playbooks for data quality, metadata/lineage, and operating models that connect data platforms to financial decisioning. Beyond delivery, Santosh serves the community as a peer reviewer of scholarly work on data/ML methodologies and as a judge/mentor for select industry and academic competitions, reinforcing peer validation and public recognition. He champions modern data culture—mentoring engineers and product leaders, and advocating automation (incl. AI agents) to elevate reliability, speed, and auditability in regulated environments. Santosh recently completed his Executive MBA, sharpening strategy and value-creation at the intersection of data, risk, and growth The meeting is over Microsoft Teams, and the joining link is https://teams.microsoft.com/l/meetup-join/19%3ameeting_NzZkYWIyOTAtODk1MC00MjVmLWJlNjUtNTRiODZmODA2Zjdh%40thread.v2/0?context=%7b%22Tid%22%3a%22bd9727e8-f539-4c76-983c-6c30130c0bee%22%2c%22Oid%22%3a%229e8d5a64-e773-4ca2-90f6-9a266129171e%22%7d See you at the meeting!Stop Guessing: Solve SQL Performance Problems with Query Store
Data Engineers in Toronto July 2026 Semimonthly Meeting Topic: Stop Guessing: Solve SQL Performance Problems with Query Store Abstract: When query performance suddenly changes, finding the root cause isn't always easy. Execution plans can change over time, and the plan cache often provides only a limited view of what happened. Query Store helps by capturing query history, execution plans, and runtime statistics, giving you the visibility needed to identify regressions, compare plans, and maintain consistent performance. In this session, you'll learn how to use Query Store to monitor query behavior, investigate performance issues, manage execution plans, and leverage built-in tuning capabilities in SQL Server and Azure SQL Database. Key topics: Query Store fundamentals Query and plan history analysis Performance regression detection Query Store reports and insights Plan forcing and plan management Automatic tuning and best practices Through live demos, you'll see how Query Store makes performance troubleshooting faster, simpler, and more reliable. Speaker: Deepthi Goguri, SQL Database Administrator Speaker Profile: Deepthi is a SQL Server Database Administrator with several years of experience in Administering SQL Servers. She is a Microsoft Data Platform MVP, Microsoft certified trainer and Microsoft certified professional with an Associate and Expert level Certification in Data Management and Analytics. Deepthi blogs for DBANuggets.com. Deepthi is a Co-Organizer for Microsoft Data and AI South Florida user group, Data TGIF, Cloud Data Driven User Group, Future Data Driven Summit, Databash Conference and Data Platform Diversity, Equity, and Inclusion Virtual Group. She is also DEI Steering Committee member for PASS Data Community Summit. She is a Redgate Community Ambassador. Along with this, Deepthi loves arts and crafts. You can contact her on Twitter @dbanuggets.Building Collaborative AI Agents Using AI Foundry Agent Service
Data Engineers in Toronto September 2026 Semimonthly Meeting Topic: Building Collaborative AI Agents Using AI Foundry Agent Service Abstract: As enterprise AI adoption grows, many real-world scenarios can no longer be solved by a single conversational agent. Instead, they require multiple specialized agents that collaborate, share context, and delegate responsibilities intelligently. This session focuses on how to design and implement multi-agent architectures using Azure AI Foundry’s Agent Service. We’ll start by breaking down when and why multi-agent systems make sense, then walk through a practical scenario where agents are assigned clear roles such as data retrieval, reasoning, and action execution. Using AI Foundry Agent Service, you’ll see how these agents are created, connected, and orchestrated to work together while maintaining context and control. Rather than focusing on theory, the session emphasizes practical design patterns, architecture decisions, and implementation considerations, including agent boundaries, communication flows, and scaling agent-based solutions safely in enterprise environments. By the end of the session, attendees will have a solid understanding of how to move from a single-agent approach to a collaborative, multi-agent model and how to apply these patterns using AI Foundry in their own solutions. Key Takeaways: - Understand when multi-agent architectures are the right choice - Learn how to design role-based agents using AI Foundry Agent Service - Explore real-world coordination and orchestration patterns - Gain practical guidance for scaling and governing multi-agent systems - Leave with a reusable architecture blueprint for enterprise AI agents Speaker: Mitul Tailor, Snr. Data & AI Engineer Speaker Profile: am a Senior Data & AI Engineer at Long View Systems, specializing in Azure services, data engineering, and AI-driven solutions. With expertise in Azure Data Factory, Fabric, Copilot Studio, and data migration, I have successfully delivered scalable and efficient data solutions. I hold a Post Graduate Certificate in Generative AI, a PG Diploma in ML & AI, and a Bachelor's in Computer Science. I am DP-600 & AI-102 certified, demonstrating my expertise in Microsoft Azure Enterprise Data Analyst solutions. Additionally, I have a strong understanding of Retrieval-Augmented Generation (RAG) AI and Azure AI Foundry, exploring how these technologies can be leveraged for intelligent automation and data-driven insights. Recognized by clients for my clear and precise development work, I am passionate about automation, AI, and robotics. Outside of work, I love to travel and have a deep interest in cars.Streaming your database - Easier said than done?
Data Engineers in Toronto September 2026 Semimonthly Meeting Topic: Streaming your database - Easier said than done? Abstract: Do you need to stream your database? So did I! Here’s what to expect. In my journey as a contractor for a major cybersecurity company, I implemented a Change Data Capture (CDC) system that aimed to seamlessly stream client databases into a centralized cloud solution for analytics and enhanced features. However, after deploying the code, I quickly realized that the reality was far more complex than anticipated. Despite using an off-the-shelf product, I encountered unexpected challenges related to scaling, data formats, integrations, and deployment—issues that took me by surprise and required significant adjustments. Over the course of 18 months, my team and I navigated these hurdles to create a robust system designed for scalability and resilience against failures. In this talk, I’ll share key lessons learned and practical tips for developers and architects embarking on similar projects. You'll gain insights into preparing for potential surprises and overcoming challenges in building a reliable CDC system. Speaker: Sigal Shaharabani, Technical Group Leader & Couchbase Ambassador Speaker Profile: Sigal Shaharabani is a Technical Leader and a Group Leader at Tikal and a Couchbase ambassador, with a great passion for backend and data systems. She started her technological career way back in 1996, working for the government, a variety of corporations, and small start-ups. Her favorite programming language is Kotlin, and will love any excuse to talk about it. In her spare time she enjoys swimming and Israeli folk dancing.
