Gartner Data & Analytics Summit in Orlando was a whirlwind of sessions, presentations, networking, buzz, and…gators! After chewing on it a bit, here are a few takeaways I hope you find valuable! TL;DR: agents on the rise, conversational everything, governance and data quality = king, and driving/demonstrating ROI and the power of people never gets old.
Written by: Chris Walker,
Head of Product Marketing at Tellius
In This Post
1. Agentic AI on the Rise
While still early days, the buzz around AI agents was palpable from Gartner experts, attendees, and vendors. Several powerful potential agentic use cases were discussed in hallways and on stage, like supply chain teams proactively adjusting plans according to impending tariffs to healthcare organizations developing assistive diagnostic tools–and everything in between.
On the ground, the reality was only ~5% of those polled onsite said they had operationalized agentic workflows or autonomous agents, which seems right considering just a handful of vendors showed video loops but practically nobody went in-depth in their demos showcasing AI agents autonomously completing complex tasks. Nonetheless, as one speaker aptly put it, “We’re not just building smarter tools; we’re creating collaborative partners that fundamentally transform how decisions are made.” And transformation doesn’t happen overnight.
2. Self-Service Analytics Isn't Dead, It's Becoming 'Invisible'
Gartner’s Rita Sallam shared how analytics has evolved from IT-led → analyst-led → end user-led (aka self-service) → AI augmented, where invisible, adaptive agentic systems that seamlessly integrate into workflows ensure trusted, responsible, AI-driven decision-making.
Rather than self-service analytics & BI being “dead,” it’s evolving into AI-driven analytics that reduces the need for manual dashboarding and SQL-based exploration. Rather than more dashboards, AI-driven analytics is embedding insights directly into workflows. Generative AI is enhancing collaboration by automating narratives and insights sharing. The focus is shifting to conversational AI interfaces and agentic analytics, where AI doesn’t just answer questions but also proactively suggests next steps and acts upon them. These themes were especially evident in this year’s Analytics & BI Bakeoff, where vendors showcased conversational interfaces and data storytelling capabilities, alongside some top-notch dad jokes like Australia’s data KOALA-ty problems 🐨 📊
3. Ubiquitous Conversational UIs
GenAI-powered conversational UIs seem to have become ubiquitous, with chat-based interfaces being the most common manifestation. However, the true potential of chat interfaces lies beyond simple question-and-answer interactions to delivering deep, business-critical insights.
To achieve this, conversational interfaces need to be intelligent and equipped with guardrails that ensure the reliability and accuracy of the insights they provide. This involves incorporating advanced natural language processing (NLP) capabilities, machine learning algorithms, and domain-specific knowledge to enable complex queries, correctly interpreted context, and provide useful/insightful responses.
Furthermore, guardrails such as data validation, error checking, and bias detection mechanisms need to be implemented to ensure the trustworthiness of conversational insights, which is crucial to avoiding costly mistakes and maintaining user confidence in the system.
4. AI-Ready Data & Observability Are Must-Haves
No Gartner D&A conference would be complete without multiple sessions related to data governance, management, and security. With the meteoric rise in all things AI (the yummy cupcake, as one session put it), governance and observability is the necessary kale salad. The main point was that while AI-ready data is necessary—consisting of high-quality, well-governed data to minimize hallucinations, enhance transparency, and build trust—governance frameworks must evolve to incorporate dynamic risk indicators and real-time monitoring.
5. Costs & ROI
Another theme that surfaced multiple times was cost containment and proving ROI on D&A investments. On this point, Gartner predicts that “through 2028, at least 50% of GenAI projects will exceed budget due to poor architectural decisions,” with the bulk of AI model costs (~70%) coming from inference rather than training, requiring businesses to rethink cost allocation. With only ~11% of organizations currently measuring the business value of governance initiatives, the expert and practitioner advice was to track not just compliance but also business outcomes and D&A’s impact on decision-making. Organizations leveraging governance impact metrics report higher executive buy-in and funding for data initiatives. Some best practices include:
✔️ Running extended pilots to validate Total Cost of Ownership (TCO) assumptions.
✔️ Automating model selection and routing to optimize for efficiency.
✔️ Implementing FinOps governance to track and control AI consumption.
6. The Human Factor Remains Key
Throughout the summit, it became evident that the biggest blockers to AI adoption are cultural, not technical, with many companies struggling with internal alignment on governance. AI adoption will come down to upskilling, AI literacy, and how seamlessly it integrates into workflows. Organizations that focus on creating trustworthy data layers and training users effectively will have the edge.
This theme resonated as technical leaders emphasized how their AI initiatives were hampered not by technology but by organizational readiness. Companies leading the way have implemented comprehensive AI literacy programs extending beyond technical teams to include business users and executives.
The most compelling case studies came from organizations that had reimagined their workflows from the ground up rather than simply inserting AI into existing processes. These companies reported significantly higher adoption rates and ROI because users experienced AI as a genuine productivity enhancer rather than just another tool to learn.
In Summary
To summarize:
1️⃣ Agentic AI is gaining momentum despite being in early adoption stages, with only about 5% of organizations having operationalized these workflows.
2️⃣ Analytics is evolving from self-service to “invisible,” AI-augmented systems that integrate seamlessly into workflows.
3️⃣ Conversational UI has become ubiquitous, though its true potential extends beyond simple Q&A interactions.
4️⃣ Creating AI-ready data with strong governance frameworks remains crucial for success, with experts emphasizing the need for real-time monitoring.
5️⃣ Cost management emerged as a significant concern, with predictions estimating that 50% of GenAI projects will exceed budgets through 2028.
6️⃣ The human factor remains paramount, as cultural challenges rather than technical limitations continue to be the biggest barrier to AI adoption, with successful organizations focusing on comprehensive AI literacy and workflow redesign.
We’re already looking forward to next year’s summit and excited for where the data & analytics industry is heading in 2025 and beyond!
Want to learn more about Tellius and see it in action? Schedule a demo now.
