The Two AI Applications Changing Product Cost Analysis
We know that AI is everywhere now, but where does it actually matter in manufacturingcost analysis?
In a recent episode of the Altium Academy podcast, our CEO Sasan Hashemi outlined two practical, transformative ways AI is reshaping how manufacturers interact with product cost data. For cost engineers facing increasingly complex decisions about materials, suppliers, and carbon footprints, these applications deliver measurable value.
Sasan identified two distinct areas where AI delivers measurable impact: solving the persistent challenge of data categorization and fundamentally changing how teams interact with cost information. Both represent significant shifts in how cost engineering teams can work more efficiently.
Defining AI in Manufacturing: It's Not All ChatGPT
The first challenge when discussing AI in any industry is establishing what we actually mean by the term. As Sasan explained:
I think it's always important to define AI, right? I think nowadays everybody, when you say AI, you think about ChatGPT and things like that. And these are really powerful models and they're LLM models. I think there's like different usages. So being that you build up algorithms based off data, I would say this is more machine learning and that's more of a self-solve problem.
This distinction matters. For cost engineering specifically, both approaches play distinct roles. Traditional machine learning excels at pattern recognition and predictive modeling. LLMs bring something different: the ability to understand context, categorize unstructured data, and interact naturally with users.
The First Application: Solving the Data Categorization Challenge
One of the persistent headaches in manufacturing is data comparability. When you're analyzing thousands of components across multiple suppliers, products, and technologies, how do you find meaningful similarities?
Sasan identified this as a core problem where AI delivers immediate value:
I think one of the biggest problems that you have in manufacturing data is categorization and comparability, right? So, hey, I have this part. Give me other parts that have the same function. And that is a very complex query in some way, right?"
Finding comparable components goes far beyond simple keyword matching. A PCBA component might be described differently across various suppliers, CAD tools, and internal systems. Traditional search requires exact matches or predefined categories, which breaks down when dealing with diverse data sources and naming conventions.
AI-powered categorization can understand functional similarities even when the terminology differs. The technology recognizes that two components serve the same purpose despite having different part numbers, coming from different suppliers, or being described in completely different ways across your systems.
At Tset, this means cost engineers can quickly identify comparable components when evaluating alternatives, conducting should-cost analyses, or benchmarking supplier quotes. The system organizes components based on what they actually do rather than just alphabetically or by part number.
The Second Application: Personalized Data Interaction
The second use case Sasan described might be even more revolutionary for how teams work with costing data daily. Traditional software operates on a predetermined principle: developers build specific views and dashboards, and users adapt their workflows to fit those fixed interfaces.
AI fundamentally changes this dynamic. Instead of building every possible view a user might want, the software can generate custom insights on demand based on natural language queries.
Sasan illustrated this with a practical example:
I can give you a complete P&L of PCBA in all the details, and you're a PCBA engineer and some of it is interesting for you and other things of it maybe you don't care about at all. I think AI in some sense completely changes that because I think the future is that there is data and the UI is basically what you wanna know. Not that UIs will go away completely, but I think you will see going forward, especially in SaaS, way more interaction with data, with fuzzy queries where people just say, tell me this, and the software just tells them and builds the UI on the spot.
This approach acknowledges that different stakeholders need different things from the same dataset. Rather than building separate dashboards for every role, AI can tailor the experience to each user's actual questions.
New team members can ask questions in their own words. Cross-functional collaboration becomes easier when each department can query shared data in ways that make sense for their specific needs. The barrier between having data and actually using it effectively gets much lower.
The Future of SaaS in Manufacturing
This conversational approach to data interaction fundamentally changes how teams can scale their cost analysis capabilities without requiring everyone to become software experts.
Sasan predicts this shift will define the next decade of software in manufacturing: "I think this is major and that will very much dominate the SaaS vertical for the next 10 years."
AI as a Practical Tool
What stands out in this discussion is the pragmatic perspective. Sasan positions AI as a tool that makes existing cost engineering work faster, more accessible, and more accurate.
The technology addresses real pain points: data categorization that currently takes hours, cross-functional communication that requires multiple handoffs, and the challenge of making complex cost models understandable to diverse stakeholders.
Want to hear more about AI, cost engineering, and the future of manufacturing analysis?
Watch the full post episode from Altium Academy:
1. How is AI used in cost engineering?
AI is used in cost engineering primarily for two applications: data categorization and personalized data interaction. AI helps categorize and compare manufacturing components based on function rather than just keywords, making it easier to find comparable parts across different suppliers and systems. It also enables users to query cost data using natural language, generating custom insights on demand instead of relying on pre-built dashboards.
2. What is the difference between AI and machine learning in product costing?
Machine learning in product costing focuses on building algorithms based on historical data for pattern recognition and predictive modeling. AI, particularly large language models (LLMs), goes beyond this by understanding context, categorizing unstructured data, and enabling natural language interaction with cost information. Both play distinct roles in modern cost analysis platforms.
3. Can AI help with manufacturing data comparability?
Yes, AI significantly improves manufacturing data comparability by understanding functional similarities between components even when they're described differently across various systems. Instead of requiring exact keyword matches, AI-powered categorization recognizes that two parts serve the same purpose despite having different part numbers, supplier descriptions, or naming conventions.
4. How will AI change cost analysis software in the future?
AI will shift cost analysis software from static, pre-built dashboards to dynamic, query-driven interfaces. Users will be able to ask questions in their own words and receive customized views of cost data tailored to their specific needs. This conversational approach to data interaction will make cost analysis more accessible across different departments without requiring everyone to become software experts.