BookMyShow has moved analytics beyond its central data team, using Databricks Genie and Unity Catalog to let business teams query governed enterprise data independently.
Manisha Sharma29 Sep 2026
15:38IST
New Update
BookMyShow has cut data engineering tickets by 90% after deploying more than 80 Databricks Genie Agents across its business teams, shifting much of its day-to-day analytics from a central data team to self-service.
Teams across marketing, finance, live entertainment and cinemas can now ask questions about business data in natural language instead of depending on engineers to write SQL queries, build dashboards or generate reports.
The shift follows BookMyShow’s migration to the Databricks Data + AI Platform and adoption of Unity Catalog and Genie. The company has built dedicated Genie Agents for different business functions, while more advanced users can use Genie Code for independent analysis.
The impact extends beyond internal reporting. BookMyShow is using the platform to track more than 3,500 offers daily, while it says publisher data integrations that previously took months can now be completed in about a day.
“Genie has taken data access at BookMyShow from something only a handful of engineers could do to something any teammate across the company can do themselves in seconds,” said Noel Curtis, Chief Technology Officer at BookMyShow.
Analytics Was Stuck With The Data Team
BookMyShow sells around 22 million tickets a month, or 264 million annually, to more than 100 million monthly active users across movies, live experiences, sports, and events.
But its analytics infrastructure had been fragmented. The company operated a data warehouse on Amazon Redshift alongside a data lake using Amazon S3 and Athena. AWS Glue handled ETL, while Apache Iceberg served as the underlying data format.
There was no unified data platform or common catalog and governance layer. That meant analytics requests from marketing, executives, or individual business units generally had to pass through the central data analytics and engineering team.
“There was practically no self-serve capability,” Curtis said. Users familiar with SQL had some ability to query Redshift or Athena directly, but that remained out of reach for much of the organization.
The issue was therefore not a lack of data. It was how quickly people outside the data team could actually use it.
BookMyShow Builds 80+ Genie Agents
BookMyShow subsequently moved its warehouse and lakehouse workloads to Databricks and adopted Unity Catalog as its governance layer. But instead of deploying one general-purpose AI assistant across the company, it created domain-specific Genie Agents.
More than 80 agents have now been built for different business functions, including marketing, customer lifecycle <a href="https://bitcomme.com/inside-brian-schottenheimers-inept-clock-management/” title=”Inside Brian Schottenheimer's Inept Clock Management”>management (CLM), digital marketing, live events, cinemas, and movie intelligence. The approach gives each agent the context of the business function it serves.
“We find that the outcomes are much better when you’re doing that,” Curtis said.
Genie One remains available for broader company-wide questions, while teams can use their dedicated agents for domain-specific analysis. More advanced users can move to Genie Code for deeper analysis and dashboard development. The tools have now expanded across finance, live entertainment, movies and cinemas, marketing, sponsorships, and brand intellectual properties, with engineering also using the platform heavily.
3,500 Offers Become A Daily Data Problem
One example shows why domain-specific analytics matters at BookMyShow’s scale. Its marketing and CLM teams need to manage more than 3,500 offers across payment promotions, promo codes, third-party partnerships, and loyalty programs.
The dedicated Genie Agent allows the team to analyze those offers directly instead of repeatedly turning to data engineers. Business users can also ask questions ranging from tickets sold for a movie and unique transactors to gross merchandise value by city, event performance, and week-over-week changes in particular offer categories.
This shifts the role of the central data team. Rather than spending much of its time responding to individual reporting requests, the team can focus more on maintaining data pipelines, Unity Catalog, and the underlying data architecture.
Curtis said productivity has increased “by an order of magnitude.”
Governance Follows Every Query
Giving more employees direct access to enterprise data creates another problem: access cannot become less controlled simply because querying becomes easier. BookMyShow is using Unity Catalog to apply access controls, classification, and lineage to the data available through its analytics environment, particularly where personally identifiable information is involved.
The company is also preparing its data architecture around requirements under India’s Digital Personal Data Protection (DPDP) Act.
“Something that you get in the data lake ultimately percolates all the way into Genie,” Curtis said.
The objective is to ensure that the permissions applied to underlying datasets continue to govern what employees can retrieve through natural-language questions. That governance layer becomes increasingly important as AI lowers the technical barrier between employees and enterprise data.
Partner Integrations Drop From Months To A Day
The changes are also extending beyond internal analytics. BookMyShow previously needed custom engineering work when sharing data with publisher partners that list its movies and events for aggregation and lead generation. After standardizing publisher data access through OpenSharing, the company says new integrations largely involve sending an API contract and configuring the required filters.
According to BookMyShow, that has reduced integration time from months to about a day. The company has also changed how it provides analytics to Indian Premier League ticketing partners.
Instead of preparing individual post-match reports, BookMyShow built a self-service application using Databricks Apps. A single data analyst developed and deployed its front and back ends within a few hours, according to the company, before access expanded from one franchise to four within days.
From Data Gatekeeper To Data Backbone
The more significant change at BookMyShow may not be the number of AI agents it has deployed, but what those agents are doing to the traditional relationship between business and data teams. A central engineering team no longer needs to sit between every business question and the underlying data. But removing that bottleneck does not remove the need for data engineering. It changes where engineering effort goes.
BookMyShow’s central team is increasingly concentrating on the foundations: governed datasets, pipelines, architecture, and access controls. Business users can then perform more of the analysis themselves. The company expects the same data and AI infrastructure to expand into areas including real-time personalization, fraud and bot detection during high-demand ticket sales, natural-language customer support, and attribution for sponsorship and advertising.
aibookmyshowEnterpriseDatabricks
