Generative AI – Spotfire and other D&A software tools
The Gartner Data & Analytics Summit 2024 was all about generative AI, collective intelligence, AI agents. AI and humans collaborating on data management and analytics. It was an exciting event with many surprises. The annual BI Bakeoff—my favorite session—featured three vendors that all showed some form of generative AI or copilot tool. Oracle’s final section on futures included an avatar of the session chair, Aura Pope, interpreting an analysis in her AI voice and a rhyming talk track!
Read more to hear about my top takeaways from the summit and how Spotfire stacks up with the generative AI featured at Gartner.
Generative AI on the rise
Generative AI, specifically augmented assistants and copilots, help BI, Analytics, and Data Science software produce what it is wired to do. The AI agents help Power BI users create dashboards and reports, especially in Azure, and help Spotfire users create Spotfire® Apps—Industry-focused Applications. These copilots help users interrogate data, and interpret analyses as narratives. The Spotfire copilot helps with both visualization and code generation—providing real insights into the problem at hand.
Across the board, I noticed the user experience in the Visual Analytics and BI products is somewhat similar—featuring a panel on the right for chatting with the AI agent. The underlying frameworks are also similar with a RAG structure, an LLM, and an orchestrator. Some AI engines also include a machine learning engine that runs in the background to identify patterns and insights.
Michael O’Connell, chief analytics officer at Spotfire, enjoying the Gartner Data & Analytics Summit 2024
With such a focus on generative AI at Gartner, it was great to compare the Spotfire Copilot™ with other copilots on the market.
The Spotfire Copilot tool leverages several LLMs, including OpenAi GPT models and Google’s Gemini models, tailored to specific use cases and scenarios. For instance, GPT3.5 is utilized for understanding the user’s intent. Additionally, other AI models, such as GPT4 or Gemini-Pro are used for custom conversations, Spotfire Q&A, Data Interrogation, Chart Generation, Code Generation, and General Q&A. Furthermore, Spotfire incorporates augmented AI features like the Recommendations engine, which continuously assesses variable relationships in the background and suggests suitable visualizations for analysis. This functionality is akin to the AWS Q engine and Tableau Pulse.
Multimodal agents
While 2023 featured much work on LLMs, 2024 is the year of multimodal. With multimodal, AI-agent applications are combining text with images, videos, and sounds as inputs and outputs. The pharmaceutical and healthcare industries are leveraging omics into precision health, digital clinical trials, and remote care products.
Spotfire, Power BI, and Tableau are text-in and visuals-out with the promise of consumers becoming authors down the road. Analytics and software engineering are starting to morph—for example, with the Spotfire Copilot tool generating python/R scripts for data science functions that can run on an interactive user marking.
In short, we are heading to a future of data and analytics “agency” with generative AI agents replacing at least some of the steps in the access, prep, model, and dashboard process.
Presentation slide from “Top Trends in AI” by Frances Kramouzis
Threats to success
What’s the greatest threat to generative AI? Costs.
Gartner highlighted costs as a new team sport and the greatest threat to the ongoing success of generative AI—including tokenization and hyperscaler cost and energy. In his presentation on Finops, Gartner analyst Rick Greenwald noted, “One of the most transformative cultural impacts that a data and analytics leader can have on their org is transparent reporting and attribution of cost and cloud spend.”
Michael O’Connell and Rita Sallam at the Gartner Data & Analytics Summit 2024
While costs are a threat to AI, businesses are still investing more than ever. The “Top Trends in AI” presentation by Frances Karamouzis, distinguished VP analyst at Gartner, noted that there is a high demand and investment in the growth of AI across the board, with a majority of businesses increasing their investment in AI across business functions.
There is a seismic shift from experiment in 2023 to execution in 2024, with cost optimization, customer experience, and growth in new products as targets.
Who is responsible for AI?
Another key theme of the conference was the rise of the CDAO as responsible for value creation via AI and behavior changes. Many noted tech leaders were responsible for AI, data, software, costs, and cyber safeguards—increasing the responsibility of analytics divisions.
Presentation slide from “Build High-Business-Impact Habits of Top Data, Analytics, and AI Leaders” by Rita Sallam
There were many discussions on how to prioritize AI projects and the best practices in their delivery. Rita Sallam, distinguished VP analyst at Gartner, discussed six habits for data and analytics leaders to cultivate as they scale their ecosystems (shown above).
Nate Novosel, VP analyst at Gartner, noted in the presentation “CDAO Agenda 2024: Reinvent or Become Irrelevant,” from a survey of CDAO leaders, 53 percent are committed to generative AI, 50 percent have deployed data products and self-service analytics, and 41 percent are committed to deploying a D&D innovation center.
The Spotfire booth at the Gartner Data & Analytics Summit 2024, featuring from left to right: Ahmad Fattahi, Adam Faskowitz, and Michael O’Connell.
The presentation “10 Best Practices for Scaling Generative AI Across the Enterprise” by Arun Chandrasekaran, distinguished VP analyst at Gartner, made a lot of sense to me as we are in the middle of this at Spotfire with our customers. Companies today want to know more about choosing use cases, implementing robust data engineering processes, uplifting AI, and gaining more data and visual analytics literacy.
The future of AI and generative AI
The summit wasn’t all discussions of costs, threats, and surveys. There were some fun sessions on AI futures at the show. “Maverick: Gartner’s Most Provocative D&A Predictions” session, presented by Frank Buytendijk, distinguished VP analyst at Gartner, included many futuristic predictions:
- The BoardBot
- Predicting 1 in 4 companies will include a BoardBot as part of the executive decision process by 2030, handling complexity and eliminating self-interest in decision making
- A Hong Kong VC has already appointed an AI bot to their board
- Data processing in space
- Predicting that more data processing will take place in space than on Earth by 2040
- Space flight costs have fallen 70 percent over the past 15 years, and solar energy costs have fallen 80 percent over the last 10 years
- Massive orbital solar power generation is becoming possible—shifting demand closer to the energy source—only held back by infrastructure operations
- DNA data storage
- Predicting that this will be mandated to make processed foods traceable by 2035
- A weird example was provided where the band Massive Attack released a 20th-anniversary edition of its album Mezzanine as a spray can containing ~1 million copies of the original album stored as synthetic DNA
Kate Darling at the summit presenting her favorite childhood robot, The Pleo.
Kate Darling, leading expert in social robotics and MIT Media Lab research scientist, gave a joyous talk on robots called “The Future of Human-Robot Interaction.” She showed progression over the past 20 years, including the latest from Boston Dynamics. The presentation featured human reactions to human-form machines like factory workers warming up for a day’s work, with exercises in sync with the machines. She also discussed her attachment to many forms of robots—especially her childhood favorite, the Pleo.
The interactions between people and machines are remarkable. Once people give a physical machine a name and interact with it as it does things with and for you…it becomes a friend.
And it wouldn’t be a true data and analytics conference without collecting socks from the vendor exhibits… My “best socks” award goes to Gurobi and Clear Query—a dead heat with two very different sock designs – for the board room and pickleball court respectively. 🙂
Collection of data and analytics branded socks from the Gartner Data & Analytics Summit 2024.
Experience generative AI within Spotfire
As demonstrated at Gartner, interest in generative AI is not going away any time soon. Be sure to supplement your analytics journey with an AI agent that can “spot” the “fire” in your data. Check out our Spotfire Copilot AI tool, a natural language extension to Spotfire, that leverages large language models to augment business intelligence and artificial intelligence, all within Spotfire.
Ready to get started? You can download the Spotfire Copilot™ App on the Spotfire community. To stay updated about the future of generative AI and analytics, request a hands-on session and stay up with my LinkedIn feed.
References
- Top Trends in AI—Frances Kramouzis
- Top Trends in BI and Analytics—Rita Sallam
- CDAO Agenda 2024: Reinvent or Become irrelevant—Nate Novosel
- Build High-Business-Impact Habits of Top Data, Analytics and AI Leaders—Rita Sallam
- 10 Best Practices for Scaling Generative AI Across the Enterprise—Arun Chandrasekaran
- Maverick: Gartner’s Most Provocative D&A Predictions—Frank Buytendijk
- Robots—Kate Darling