ThoughtSpot becomes a sage at Beyond 2023 conference

Enterprise data analytics and AI specialist ThoughtSpot is gearing up for its Beyond 2023 conference, which takes place on 9-10 May. It claims 10,000 registrations for the online event: roughly 1,000 for every year the $4.2 billion company has been in business.

When diginomica caught up with ThoughtSpot in March, it had just launched GPT integration, allowing even non-expert employees to seek insights from corporate data using natural-language prompts. A space also occupied by the likes of Clari, Sisense, and others.

But today, it is launching a range of new products and services, with ThoughtSpot Sage at the epicentre – an enterprise data search capability that combines foundational language models with the company’s own proprietary search technology. The system will be available in private preview from today.

Also launching at Beyond 23 are:

• Integrations with common workplace productivity tools and office suites, extending the ‘platformization’ of the company

• Liveboard collaboration and interactive data features (ditto)

• A series of new database connectors (ditto)

• A new Data Modelling Studio, the latest feature in the Data Workspace (ditto)

• And the availability of ThoughtSpot SaaS on the Google Cloud (Google features prominently in today’s announcements).

Plus, ThoughtSpot Monitor for Mobile, a new AI-infused system that will proactively alert users of significant changes in key performance indicators (KPIs). 

Sumeet Arora is ThoughtSpot’s Chief Development Officer. He says:

The founding thesis of ThoughtSpot was improving human behaviour around the analysis of data, and how we ask questions. And we built a system to achieve that. But what the availability of large language models [LLMs] has done is help us put the last mile in place, if you like. To take pure natural language and translate it into data questions.

Beyond 23 is all about the modern data experience. And ThoughtSpot Sage is a big component of that. It means, for the modern data stack – which is data in the cloud – being able to help every knowledge worker access data using natural language. And nothing less. 

So, whether you are an English major, a Humanities expert, a data expert, or an engineer, it doesn't matter. Everybody can leverage the power of data and analytics using natural language. So, ThoughtSpot Sage is the natural language search experience powered by AI.

Leaving aside the question of whether using the name ‘Sage’ for a cloud-based enterprise tool is wise, given the prominence of an ERP company with the same name, he continues: 

The second pillar of our announcement is being able to collaborate on liveboards on answers, to talk to each other, make decisions, and take actions right there in the product. 

And the third is around how, when we do jobs, we are all goal driven. We have KPIs, and we like to stay on top of those goals. So, when you're on the go, personalized to you and relevant to you, we keep you on track of your key metrics, which are all powered by analytics and data.

He adds:

We are not living in a walled garden built by each of the cloud players. By working with the best of the big players, we allow our customers to get best-of-breed technology.

Reaching the common knowledge worker

So, what is the underlying technology of ThoughtSpot Sage? According to Arora, the company’s own proprietary search tool has been combined with OpenAI’s GPT engine for the natural-language element. 

But how does that play with Google and its rival Bard system? ThoughtSpot seems keen to talk up the Google relationship and the integrations it offers, so would Google prefer the company to use Bard? 

The answer is that ThoughtSpot Sage won’t be limited to a single LLM in future, says Arora:

We are working with alternate solutions and Google is one of them. We are absolutely open to them. We will provide that choice to our customers. We will use multiple systems for sure.

(Google, of course, has been in the news this month with the departure of its AI guru Geoffrey Hinton, citing concerns about the technology – something of an Oppenheimer moment for the industry.)

When diginomica last spoke to ThoughtSpot, we asked co-founder and CTO Amit Prakash whether the democratization of data analysis via generative AI parked a tank on data scientists’ lawns. Prakash said:

I can certainly see that, in different industries, this kind of technology can have a negative impact on professionals. But I'm not worried about that happening in the data space. 

If you look at the current state of the industry, there's still a lot more demand for data professionals, despite the economic downturn, than there is supply. So, most companies are not able to take full advantage of the data they have, because they can’t bring in that data in a shape that the business can benefit from. So, they make decisions based on their gut.

What we're trying to do is create the conditions where someone says, ‘I can get the answer right away, I don't need to wait for a week.’ And that makes data a lot more valuable to the company.

So, what is Arora’s take on this topic? Are companies like ThoughtSpot undermining data professionals’ expertise and value to the enterprise? Quite the reverse, he tells me:

If you look at data scientists, their work has not reached the common knowledge worker completely. So, how do you bring that power to the knowledge worker, without that worker worrying about the details of the machine learning world? 

The natural-language evolution of analytics makes it easier for data scientists to leverage their time, so their impact is going to increase. It’s the same thing with data analysts. The data modelling layer and the business language need to come together. And AI will help with that, but it will require governance and expertise from those analysts. 

In other words, the analysts will be the ones who will bring AI to enterprises’ private data. So, I'm optimistic, because data analysts have long lived in a land of drudgery, building reports and fixing reports. But now they are going to be able to do what they were employed to do, which is focus on business outcomes: reducing fraud, reducing churn, and improving the value chain for customers.

Data needs to be near

Something else that Prakash talked about in March was that, using these solutions, there is less need for the kind of data categorization and management favoured by the likes of Informatica. Just throw your data at large-language-powered tools and see what insights stick, he suggested. What does Arora think? He says: 

We need both. The reality of the world we live in is that there is a place for high-quality, curated data with the right data models. But let's also recognize that our world is moving very fast. And it takes time to build high-quality, governed data sets and data models. 

If you don't analyze your data quickly, it becomes old just as fast. So, there is a place for that too, actively solving problems with natural language.

Some of ThoughtSpot’s competitors favour a niche, task-specific approach, with the likes of Clari and Sisense also turning to natural-language queries and GPT-enhanced processes. By contrast, ThoughtSpot seems to have a more generalist strategy. Arora concurs with that assessment and says: 

Yes, ThoughtSpot is a platform, one that enables a modern data experience that is natural language based. We want people to ask their questions and give them insights in return: what happened? Why did it happen? What is likely to happen? And what can I do to change what is likely to happen? We want to cover that spectrum in a natural language format. 

More often than not, silos don't work. Data needs to be near, because when you are working in business, the entire value chain is connected. HR cannot be HR by itself, and manufacturing cannot be manufacturing by itself. People are important, and the supply chain is important, and the factory is important. 

So, by having this horizontal play, we have no limits on curiosity, and there are no limits in what you can bring together to analyze.

Only the paranoid survive

Microsoft is a huge investor in ThoughtSpot partner OpenAI and its GPT engine. Does Arora worry that Microsoft may be muscling in on ThoughtSpot’s territory, building the same capabilities into its cloud-based suites?

He says:

I'm both excited and paranoid. ThoughtSpot’s founding thesis was a platform that matches human behaviour. Now, if Microsoft is copying that, what better validation do you want? When the second largest company by market capitalization in the world wants to take the same approach that we took? That's good news for us. I love it.

In March, Microsoft announced that Microsoft Search combined Azure OpenAI and Cognitive Search to “revolutionize your data with ChatGPT” within Microsoft 365. There is no suggestion this was copying ThoughtSpot’s model, merely that it was the obvious thing to do.

Arora continues:

But I am more paranoid about the barrier to entry in technology being so low that, any graduate with a fresh approach to solving this might have ambitions too. So, inside ThoughtSpot, there is blue-sky thinking as well, about how we are evolving our platform.

GPT alone is absolutely not enough; it has a lot of errors. So, the system we have built leverages our relational search technology and puts the human in the loop. It satisfies security, trust, accuracy, and relevance.

GPT and LLMs are one component of the entire system that is needed to solve the analytics problem. So, we are excited about that, but at the same time we are always exercising our brains. Because only the paranoid survive.

Finally, is he worried about data stress and burnout? With ThoughtSpot Monitor for Mobile, a system that issues mobile alerts on corporate KPIs would seem to be edging us towards never being able to leave the office, as we react to whatever machines tell us 24 hours a day.

Picture the constant state of anxiety as we measure success over shorter and shorter timescales. “Shares are down this afternoon. Quick, do something!” A world in which those machines, perhaps, are running the enterprise: the corporate equivalent of a fitness tracker that orders you to run around the park. 

He says: 

No, I'm not concerned about that. The biggest challenge in business is to avoid noise. So, you will only be alerted when you should be alerted. And you'll be thankful, because otherwise you'd be spending your time worrying about it – digging into systems every day and asking, ‘Am I on track? Or am I not?’ 

So, it will actually save you time and money. That's the problem we're solving.

My take

The platformization of data analytics continues apace, suggesting that ThoughtSpot was in the right space at the right time. But companies with much bigger platforms are watching.