Putting Data to Work Requires Someone to Build for Everyone
Every organization says it wants self-service analytics, but very few have it.
What actually stalls true self-service analytics is the invisible work that has to happen before anyone can ask a data question: the semantic layer, query engine, security model, and analytics app.
That work lands on your product and data teams, who are already time-strapped to deliver core features and strategy, and it lands on them again every time a new question, a new tenant, or a new dashboard request arrives.
So the backlog grows, and business users wait. The promise of putting data to work across the business stays a slide in someone's strategy deck.
The real bottleneck comes down to two problems: the cost of building and maintaining that analytics layer as definitions and context change, and the gap between a business question and an answer someone can actually trust.
ThoughtSpot Is Now Live in the ChatGPT Work Plugin Directory
ThoughtSpot is now available in the ChatGPT Work Plugin Directory, for both halves of that problem.
SpotterCode is your AI Coding Partner: Turning instructions into execution, so the analytics layer stops being a multi-sprint project.
Spotter is your AI Analyst: Giving every business user governed, verifiable answers to any question, grounded in relevant data and context, and delivered with recommendations that anyone can act on.
One builds. One answers. Both live in the conversations your teams are already having.
“Analytics should work where your teams work. With the Data agent in ChatGPT Work, teams can build ThoughtSpot AI analytics in minutes using natural language. Business users get governed answers based on definitions they already trust, while developers ship working embedded analytics apps from a single prompt.”
Bhargav Addala
SVP, Product Management, ThoughtSpot
“We’re excited to see more analytics partners bring their capabilities to ChatGPT Work, giving business teams more ways to ask questions in natural language, build dashboards, and put their data to work.”
Arpan Shah
General Manager, Technology Vertical, OpenAI
SpotterCode: Turn Instructions Into Execution
SpotterCode builds and acts on your live ThoughtSpot instance. It goes well beyond a documentation search feature or a code snippet generator.
Describe what you're building in natural language, and SpotterCode:
Ships analytics that run, not something to finish: Spotter and Liveboard components come with the Visual Embed SDK configured, auth wired, and data models connected.
Handles the plumbing automatically: CORS, security settings, and authentication work correctly out of the gate.
Acts on your live ThoughtSpot instance: Tenant creation, user management, and more happen with automated REST API calls, with users describing the outcome instead of just clicking
Knows your platform, not a generic version of it: Grounded in ThoughtSpot's live developer and API documentation, so it’s not guessing at SDK behavior.
However, what matters is what your team does with the time it gets back, not just the code.
Leveraging ThoughtSpot, Thrive Learning embedded AI agents for more than 56,000 active users in just three days and has cut analytics support tickets by 81% since.
Endpoint Clinical took just three months to go from zero to launch, with a team of three fractional FTEs completing the full integration and data modelling pipeline. This lowered upfront development costs and freed the broader product team to stay focused on mission-critical features.
Spotter: Governed, Accurate Answers for Everyone
Fast to build is only half of it. What comes out has to be transparent, accurate, and relevant.
Spotter brings ThoughtSpot's Agentic Analytics into the Data Agent in ChatGPT Work, so anyone in your organization can:
Ask any question in plain language: No query syntax, no dashboard hunting, no waiting on an analyst's queue. Spotter handles even the most complex of business questions.
Check the answer before acting on it: Every response resolves against your governed metrics and shows its work: the logic in search tokens, traced back to approved definitions.
See what changed, what’s next, and why: Spotter investigates the drivers behind a metric change, forecasts where it's headed, and lets you drill into any answer with follow-up questions.
Make decisions with the full picture of your data: Spotter natively connects to structured and unstructured data within B2B apps your teams already run.
Every answer is grounded in ThoughtSpot's semantic and context layer, meaning that reported "revenue" aligns with your finance team's definition, rather than an LLM's interpretation of a column name.
With Spotter, Ecolab eliminated manual month-end reporting processes that previously took days and reduced complex billing analysis from months to weeks.
They also gained executive adoption and support through improved visibility and created a foundation for scaling analytics across global operations.
As Anand Iyer, Chief Data Officer of Ecolab, stated, "[t]he game changer was the ability to combine computational AI with some of these self-serve analytics tools where people could just ask English-language questions, get data they could then use, and [then] save those queries."
Trusted by Design, Not by Exception
Moving analytics into a chat interface raises an obvious question: what happens to governance? Here's our answer.
With ThoughtSpot and the Data agent in ChatGPT Work, your existing controls carry through unchanged:
Row-level, column-level, and attribute-based access controls apply exactly as they do in ThoughtSpot. Users see what they're authorized to see, and nothing more.
Your identity provider stays your identity provider, with SSO, SAML, and OIDC federation and per-user security enforcement.
Every connector is permissioned individually, so you decide case by case what an agent can act on.
Full audit logging, including actions initiated by AI agents, retrievable via the Audit Log API or pushed to your SIEM.
No data copied out, and no prompt persistence. Questions are answered against your governed data where it already lives.
This is the same enforcement layer that already runs your analytics, simply extended to a new surface in ChatGPT Work.
How to Build What Differentiates You
Agentic AI changed what your team can build. It didn't change what your team should build.
The winners in this next phase won't be the ones who rebuilt analytics: a semantic layer, a query engine, and a security model that were already solved.
They'll be the ones who spent that time on what only they could build, and let governed analytics show up wherever their people already work.
Both plugins connect to your ThoughtSpot instance. Find Spotter and SpotterCode in the ChatGPT Work Plugin Directory.
Don’t have an instance yet? Request a demo or try ThoughtSpot free today.
