Many years ago, I developed applications for AutoCAD to support product development process and analyze data and customer feedback in engineering and architecture firms. What was always fascinating is to see how powerful was AutoCAD data model allowed to customize data sets and create a variety of engineering solutions, later supporting query and displaying this data for decision making. Any query engineers needed to do were translated into queries selecting data, later on RDBMs support for AutoCAD was introduced allowed to connect and store data in databases. We were able to store data in relational tables and query it using SQL to product results and data analysis. It was part of product data management.
For more than two decades of me working on different engineering solutions, main data management challenges in data management systems was about how to figure some sort of query language that needs to be run to get data – it was AutoLIST, SQL and later variety of other data processing mechanisms. Variety of NoSQL databases, search indexes, RDF/OWL storage, Graph Databases – all of them required some technical understanding of the data to get those queries and show the results to customers.
In my view, what was the key in all these solutions and usually differentiated more successful systems from less successful was. how efficiently system was able to help users and engineers writing queries to get information out and show it to users. The last resort for all PLM application engineers and creators of solutions was always SQL language, which allowed to query multiple PLM and other databases. As data becomes more intricate and highly diversified, understanding its structure, relationships, and context is crucial. Making the right query and solution is not only about running a query, but also about comprehending the bigger picture of product data. Moreover, modern data management solutions are escaping from RDBMS paradigms and SQL language and brining the need to use variety of data query end points, languages and APIs.
The Rise of AI in Data Management and Queries
We are now witnessing a surge of new tools that promise to redefine how we interact with data. And all enterprise software vendors are looking on this process with slight shock – new tech allows to combine data about raw materials, manufacturing process, supply chain management, product lifecycle management, project management disciplines together.
One of the most remarkable developments is the shift towards AI-driven query generation. The next dominant “programming language” may not be Python, Java, or SQL—it might be plain English. With the advent of generative AI models, human-machine interactions are rapidly evolving, making natural language the new interface for database queries.
A recent example that caught my attention is QueryGPT by Uber, an AI-powered tool that translates natural language into SQL queries. Instead of manually writing complex queries, a user can simply ask, “How many trips were completed by Teslas in Seattle yesterday?” and receive an optimized SQL query instantly. No coding required.
Is Query Programming Dead?
For decades, SQL has been an essential skill for data professionals. Later, the landscape of query languages was extending to use various query languages and APIs and frameworks developed by different database tools. AI-driven tools are automating much of the manual query work, raising a key question: What is the future of product data queuing in modern PLM software tools?
Here’s my take:
- Query Languages aren’t dead—they are evolving. AI can generate SQL queries (or other API and queries), but understanding data structures, optimization, and debugging will still require human expertise. The ability to manage and interpret data remains invaluable.
- AI is a co-pilot, not a replacement. Just as calculators didn’t kill mathematics, AI won’t eliminate database programming and query languages—it will redefine how we interact with data. Engineers and data professionals will shift from writing queries to refining AI-generated queries and validating results.
- Knowing how to work with AI-powered query tools is the real skill. Instead of memorizing query syntax, the future belongs to those who can effectively prompt AI models, verify outputs, and ensure data accuracy. Understanding the data model behind the AI-generated queries will remain essential.
- Future query interface in the modern PLM software might be redefined and to be AI driven. Instead of learning specific queries for Teamcenter, Windchill, Aras or other query PLM systems. the future will provide us a way to run queries more automatically. Those who can effectively prompt AI models will be wining PLM implementation race and those PLM systems that provide such interfaces will lead.
AI-Powered Querying of Complex Product Models
This transformation leads me to an important question: What about product data? The data that is becoming super complex, intertwining requirements with engineering, manufacturing and suppliers data?
I want to think about a future where AI-driven tools allow a procurement manager to query complex product data models—perhaps even entire product knowledge graphs—using natural language. Instead of manually navigating through intricate PLM/ERP User Interfaces, a user might simply ask, “Show me all alternate suppliers for this part with lead times under two weeks and used in our products before.” The system would instantly generate the relevant query, pulling data from product knowledge graph and multiple sources.
The New Role: Data Engineers?
As PLM systems integrate AI-driven query capabilities, they will still require engineers to understand data logic, relationships, and semantics. The complexity of product data—BOMs, revisions, manufacturing constraints, and regulatory requirements—demands an intelligent approach. AI will help, but engineers will need to master the logic behind their data structures.
Looking ahead, engineering and manufacturing software will increasingly abstract query complexity, hiding SQL, Cypher, Search (or whatever query language comes next) behind AI interfaces. However, engineers will still need to understand the logic and meaning behind their data.
Thinking about new ways of query data, reminded me of how engineers worked with AutoCAD 30 years ago often wrote custom scripts to extend its functionality. Similarly, engineers of the future might not write data queries, but they will need to “program” AI models—structuring prompts, validating outputs, and understanding the underlying data logic.
What is my conclusion?
AI is transforming database work, making complex data queries more accessible. But while AI tools will generate queries, PLM tools will have to build an infrastructure and data models to make layers of data available to make these systems to work. It starts from engineers understanding their data. The shift isn’t about eliminating SQL or other query languages — it’s about evolving from query writing to data logic mastery. As AI takes over syntax and execution, the real skill will be knowing how to work with AI-powered systems to extract meaningful, accurate, and reliable insights.
Just my thoughts…
PS I’d love to hear your thoughts. How do you see AI shaping the future of engineering data management?
Best, Oleg
Disclaimer: I’m the co-founder and CEO of OpenBOM, a digital-thread platform providing Collaborative Workspace with PDM, PLM, and ERP capabilities. With extensive experience in federated CAD-PDM and PLM architecture, I’m advocates for agile, open product models and cloud technologies in manufacturing. My opinion can be unintentionally biased.
