Global automotive supplier TI Automotive operates at the heart of a rapidly transforming mobility industry. The company is a market-leading Tier 1 supplier with advanced expertise in fluid management and lightweighting technologies, delivering safety- and performance-critical systems used across the automotive value chain.
As automakers accelerate electrification, digitization, and supply chain transformation, data and AI are becoming increasingly central to engineering, manufacturing, and operational decision-making. Within this environment, building effective data products and ensuring their adoption across the business has become a strategic priority.
In this final installment, Apurva Wadodkar, Senior Director and Head of Data and AI at TI Automotive, joins Merav Yuravlivker, Chief Learning Officer at Data Society, to discuss how the organization is driving adoption of data products through what she calls the “Data Buffet” framework, along with practical strategies for upskilling teams and preparing for emerging data challenges in manufacturing.

Part 1 of this series explained what it takes to build an AI practice from the ground up through education, prioritization, and disciplined execution.

Part 2 examined two foundational capabilities that determine whether AI initiatives scale successfully across the enterprise: AI governance and decision intelligence.

The “Data Buffet”: Organizing data products around business domains

For Wadodkar, the success of a data organization should not be measured by how many models or pipelines reach production.

The real metric is adoption. “We are not just executors, we are partners,” says Wadodkar. “We don’t rejoice when data products go into production. That’s just a milestone. The real milestone is adoption.”

To drive that adoption, Wadodkar introduced a framework she calls the “Data Buffet.” The concept reframes how data products are organized and consumed within the organization.

“It is like a buffet you would go to, where you have different cuisines,” she explains. “In our case, those cuisines are domains like finance or supply chain. All the data products related to that domain are bundled together so people know exactly where to go.”

How the Data Buffet works

The framework organizes data products in a way that mirrors a curated dining experience:

• Cuisines, or business domains. Examples include finance, supply chain, or operations.
• Menu, or available data products. Each domain defines which data assets are relevant to the organization.
• Kitchen, or data engineering layers. The bronze and silver layers act as the preparation stage before products reach the buffet.
• Cooks, or data engineers and product teams. They ensure the underlying ingredients (data) are reliable and well-sourced.

But the biggest shift in the model is who designs the menu. Traditionally, data teams define the available datasets and dashboards. Wadodkar’s approach brings business stakeholders directly into the design process.

“We are bringing the business into the kitchen,” she says. “Finance, for example, helps create the finance menu alongside engineers and product managers.”

This collaborative structure creates domain pods, where technical teams and business leaders jointly determine which data products the organization actually needs.

The result is a system where business teams are not just consumers but co-creators of the data products they rely on.

“Because they helped create it, adoption becomes natural,” Wadodkar explains. “If I helped design the menu, I’m going to eat from it.”

Fixing the “Leaky Faucet” problem in data pipelines

A critical part of making the Data Buffet successful is ensuring the underlying data is reliable. For Wadodkar, that means going beyond traditional data engineering practices and addressing issues directly at the source.

“A data organization’s role doesn’t just start at the bronze layer where we extract data,” she says. “It goes all the way back to the source systems.”

Too often, teams attempt to fix downstream issues without addressing upstream problems in data entry or business processes. Wadodkar refers to these recurring data issues as “leaky faucets.”

“You can keep standing there with a broom trying to clean the water, or you can fix the faucet,” she explains. Her approach encourages data teams to actively identify these upstream problems and escalate them to the appropriate stakeholders so they can be corrected at the root.

This mindset shifts the role of data leaders from pipeline builders to process improvers, strengthening data quality across the organization.

Upskilling teams by embedding standards into architecture

Upskilling data teams is often treated as a separate initiative, but Wadodkar argues it should be integrated directly into the organization’s technical architecture. “Upskilling needs to be planned alongside your yearly programs. You plan people programs just like you plan technology programs,” she says.

Beyond training programs, Wadodkar emphasizes embedding best practices directly into the design of the data platform so teams naturally follow them.

Key elements of this approach include:

1. Architectural Standardization: The “kitchen” layers of the architecture are designed to enforce repeatable patterns across data pipelines. “You don’t want to rely on memory for people to remember the architectural guidance,” she says. “You bake it into the design.”
   2. Clear data product descriptions: Just like dishes at a buffet have labels, every data product must include a clear explanation. “Any data product going into the gold layer needs a good English description,” Wadodkar says. “Don’t make it technical. It should be understandable to anyone.”
2. Consistent data structures: Standardization extends to the structure and format of data products.

Examples include:

• Consistent timestamp fields
• Binary flags represented as 1s and 0s, not multiple variations
• Uniform naming conventions across tables

“When users pick up different products from the buffet, they should see the same structure every time,” she explains.

4. Governance gates: Finally, architectural checkpoints ensure that all products entering the “gold” layer meet the required standards.  These approval gates reinforce consistency and reliability across the data ecosystem.

Tariffs will become a data problem

Looking ahead, Wadodkar sees one issue becoming a major priority for data leaders in manufacturing: tariffs. Global trade policies and shifting supply chains are creating new operational complexity for manufacturers.

“Every data leader now has to think about how we automate tariffs being put on imported goods,” she says.

Because tariffs change frequently, companies need systems capable of capturing, calculating, and distributing these costs accurately across the business.

For example, organizations may need automated workflows that:

• Track tariff costs applied to imported components
• Allocate those costs to specific products or customers
• Generate accurate billing adjustments
• Provide real-time reporting for finance and supply chain teams

“You have to build an accounting system that says we paid this much in tariffs and here’s the bill,” Wadodkar explains.

Tariffs also influence supplier strategy, prompting companies to reassess sourcing decisions. “If I’m paying this much in tariffs, maybe I should move sourcing elsewhere,” she says.

For data teams, this means supporting supply chain optimization and financial forecasting in an increasingly volatile environment.

The strategic role of data leaders in decision quality

Ultimately, Wadodkar believes the impact of data leadership extends far beyond technology. The real contribution lies in improving how organizations make decisions.

“The difference between a successful company and an unsuccessful one is the quality of decisions they make,” she says.

Data leaders sit at the center of that transformation. By ensuring reliable data, building usable products, and embedding analytics into everyday workflows, they directly influence business performance.

“As a data person, you can enhance the quality of those decisions. You are directly contributing to the company’s growth,” Wadodkar says.

Her advice to data leaders is simple but powerful:  “Think of yourself as partners. Start from there.”

CDO Magazine appreciates Apurva Wadodkar for sharing her insights with our global community.