How Rivian drives trusted, AI-powered decisions at the speed of thought with Databricks

Rivian is building electric vehicles and services that require fast, trusted decision-making across manufacturing, commercial touchpoints, service interactions, finance, people, supply chain and operational planning.

Leaders need fast access to the right insights. Analysts and business users need well-defined data and metrics they can trust. Technical teams need a scalable way to serve high-quality data products to many audiences without duplicating logic or adding operational overhead.

With AI, Rivian sees an opportunity to change how people work with data — moving from predefined reporting to self-service analytics that can support faster, more trusted decisions.

“At Rivian, we are heavily leaning on AI and tech partners to rethink analytics: full self-serve, high trust at the speed of thought,” said Romit Jadhwani, Sr. Director, Enterprise AI, Data & Productivity at Rivian. “We leverage the platform to empower our employees to be thought leaders. Databricks is a key partner in making this happen.”

With Databricks, Rivian is building a governed intelligence layer that supports dashboards, natural-language exploration, custom AI/ML data applications and AI-powered workflows from the same trusted foundation. The goal is to move beyond a limited menu of predefined cuts and reports, giving users a way to explore trusted data, ask follow-up questions, build solutions and act on insights without waiting on a centralized analytics queue.

Databricks + Rivian Partnership on AI/BI

Rivian started by partnering closely with the Databricks AI/BI product team to migrate a massive, multi-domain dashboard base to Databricks AI/BI in less than six months.

The migration was only one part of the story. Rivian worked with Databricks as a design partner, identifying and shaping roughly 58 new AI/BI product features that now benefit the broader Databricks user community.

As Rivian consolidated analytics into Databricks AI/BI, the team could keep its semantic layer, governance and sensitive data controls in one place. That includes personally identifiable information, highly restricted financial data and other sensitive business data that must remain governed, auditable and permissioned.

For Rivian, this consolidation is central to the long-term vision. By building on Databricks, Rivian can bring data, semantics, permissions, dashboards, AI and custom applications together on one governed platform.

Building trust with Metric Views and Unity Catalog

In every fast moving data enabled organization, every data user faces the same question - "Which metric do I trust?”

At Rivian, the answer is to define critical business metrics once, certify them and make them available through a governed semantic layer on Databricks. Rivian is using Unity Catalog metric views to standardize metrics with transparent logic, lineage and permissions.

The team is actively building Rivian’s company scorecard metrics into Unity Catalog metric views, with a goal of standardizing more than 50 metrics. This gives users visibility into definitions and underlying tables while preserving permissioning patterns inherited from the source data. For sensitive metrics, including financial measures, inherited governance is essential.

Rivian is also using AI to accelerate the certification process itself. As the business changes, the central data team needs to keep certified data products current across domains. Rivian’s in-house AI agent helps fast-track certification by supporting review and validation for new data sets.

That flexibility lets teams move fast without weakening trust. Certified metrics give business users confidence when they need an official number, while custom metrics give analysts and domain teams room to explore when the business is evolving.

Once the context lives in a governed semantic layer, BI becomes just one of several ways teams can work with the same trusted metrics.

“Now we are investing more in semantic layers where the context is housed, and our conversational agents, dashboards and Databricks Apps are all based on that singular place.” said Sahil Aggarwal, Senior Staff Analytics Tech Lead at Rivian and Volkswagen Group Technologies (RVTech), Rivian’s joint venture with Volkswagen. 

That is the foundation for scalable self-service: data products that stay governed and current as the business changes.

Expanding self-service with AI-powered analytics

In many legacy BI and data lake environments, self-service still depends on technical skill. Data engineers and SQL analysts can work directly with the data, while business users are often limited to existing dashboards or analyst support.

Rivian’s vision for enterprise-wide self-service is different: eliminate the skills barrier so every person can become a data person. With Databricks AI/BI, Genie, certified metrics and AI-assisted development, users can solve more of their own problems with trusted data.

That shift is already visible across Rivian. Using Databricks AI/BI, Genie Code and Databricks Apps, business users are building solutions that previously would have required deeper technical support: a finance analyst with no SQL experience built an end-to-end CFO revenue dashboard with complex data transformations, a treasury manager created an AI/BI dashboard to understand cash positions, and a supply chain analyst built a Databricks App to track inbound inventory shipments.

For Rivian, that is what true self-service looks like. Teams are developing proofs of concept and creating data applications to answer complex business questions in days instead of months.

“When your metrics are certified, governance is unified, and your AI layer can meet users in natural language, the skills barrier disappears,” said Saritha Suresh, Principal Product Manager, Enterprise Data & Analytics at Rivian. “With Databricks AI/BI, Unity Catalog, and Genie, we’ve stopped asking, ‘How do we enable business users?’ and started asking, ‘What will they build next?’ That’s the shift.”

As more business users build and explore on trusted data, analytics teams can spend less time recreating logic across dashboards or responding to one-off requests. Their work shifts toward building better certified data products, strengthening the semantic layer and expanding the AI-powered workflows that help teams answer more questions on their own.

That creates a different relationship between the business and data. When users can get trusted answers in the moment, they ask better follow-up questions, explore more context and make decisions with fewer handoffs.

“Convenience is the order of magnitude,” said Michael Flynn, Director, Big Data & AI at RVTech. “When those timescales collapse, people will actually ask the follow-up questions and get the answers from the source of truth.”

The same foundation is also helping Rivian bring AI-powered analytics into the operational workflows that matter most to the business.

Bringing AI-powered analytics into operations

Rivian’s governed analytics foundation is already supporting high-value use cases across manufacturing, supply chain and operations.

Real-time supply chain visibility cuts monitoring time by 60 to 70%

Supply chain planners previously spent hours each day checking multiple systems and compiling status updates from spreadsheets. With the unified real-time dashboard built on Databricks, teams can now monitor inbound supply, identify risks earlier and act before problems escalate.

The dashboard includes automated Slack alerts and proactive risk identification three to five days in advance. Teams have moved from reactive firefighting to early intervention, reducing monitoring time by 60 to 70%.

Unity Catalog provides the governed, trusted data foundation that makes this kind of cross-system visibility possible at scale.

Faster root cause analysis reduces dependence on tribal knowledge

Investigating an inventory discrepancy used to take more than 30 minutes of manual cross-referencing across systems. Planners often had to rely on institutional knowledge about specific suppliers, shipment patterns or operational exceptions.

Now the same investigation can take under two minutes. Parallel queries automatically surface linked inventory positions, shipment status, quality issues and production plans in a single view, with historical supplier risk context built in.

This speed is powered by Databricks ML and Unity Catalog’s ability to connect and govern data across relevant sources. Instead of searching across systems, teams can investigate from one trusted foundation.

ML-driven predictions proactively reduces inventory stock-out risks

Stock-out risks were previously difficult to track at speed. Days-on-hand calculations were manual, and risk was often identified only after thresholds had already been breached.

Now, ML models on Databricks predict inventory stock-out more than four days ahead and automatically score parts as red or yellow before they hit critical levels. This gives planners time to act before a production line is at risk of going down due to inventory.

With Databricks, Rivian can bring together governed data, ML-driven predictions and operational workflows so teams can move earlier and with more confidence.

AI agents accelerate trusted data products

Rivian is also leaning into AI to scale the work of platform and data engineering teams.

The company operates across many domains and source systems, including homegrown and third-party applications. Ingesting, certifying and maintaining trusted data products across that environment requires speed, consistency and strong governance.

By combining Databricks with AI-assisted engineering workflows, Rivian has reduced new ingestion setup time by more than 60% in some scenarios. That helps data teams move faster across new domains and sources while giving the business access to trusted data products sooner.

The result is a more scalable operating model: AI helps data teams build, certify and monitor trusted data products, while business users consume and act on those products through dashboards, Genie, Databricks Apps and AI-powered workflows.

Making AI a thought partner across the business

Rivian’s analytics strategy is moving beyond a world where only analysts and data teams can get to the right answer quickly.

The company is well on its way to full self-service at scale: finance analysts building dashboards without writing SQL, planners seeing inventory risk before it hits the line, data engineers using AI to set up ingestion faster, and business users asking follow-up questions in Genie instead of waiting for custom requests.

That only works with the right foundation. Certified metrics, governed data products and Unity Catalog give Rivian the context, permissions and lineage needed to scale AI-powered analytics without creating new versions of the truth.

For Rivian, this is the next model for enterprise analytics: AI as a thought partner for anyone making decisions with data, backed by a platform built for governance, scale and accuracy.