Like a kid outgrowing shoes, AI might be moving too fast to the point where other integrated parts cannot catch up.
While AI tools are dramatically compressing hardware design cycles and introducing new products, sourcing workflows remain slow, manual and disconnected.
This gap creates risk such as part shortages, redesigns, and delayed launches.
Leveraging tools that provide real-time component availability, pricing, risk signals, and compliance directly into the design phase can help save companies time and money, especially with the uncertainty of tariffs factors such as inflation.
I sat down with Ryan Chan, Vice President, Solutions Consulting, Supplyframe to dig a little deeper.
How does AI-driven design accelerate hardware development cycles, and where do sourcing and procurement workflows struggle to keep pace?
AI-driven design benefits hardware development cycles in a number of areas, including simulation, testing, validation, layout, and Design for Manufacturing (DFM) efforts. For example, Generative models can explore thousands of power, thermal, cost, and size tradeoffs in hours instead of weeks.
Furthermore, AI-accelerated simulation allows for faster and more efficient testing and validation. The rise of digital twin technology is a great example of this use case. AI models can also flag manufacturability risks during the design phase, which prevents redesigns or other delays later in the process.
As part of this ‘shift left’ approach that focuses on risk mitigation and sourcing viability at the point of design, teams can sidestep much of the risk associated with product designs through the use of intelligence and insight, paired with AI assistance.
In terms of where teams struggle to keep pace, sourcing and procurement remains a decidedly reactive effort. This, combined with human-driven workflows means that errors are common, and issues aren’t identified until they’ve become a problem.
A lack of real-time intelligence in a constantly changing and extremely volatile market means that teams are flying blind more often than not. Even when this data is available, it’s often fragmented, outdated and siloed, which doesn’t address the foundational issues.
To solve this, teams should look to utilise design-embedded sourcing intelligence, predictive risk models, and recommended alternates to quickly address high-risk parts in the BOM.
What risks emerge when design velocity outstrips component availability, pricing visibility, and compliance checks?
A design that only takes into account engineering risks can unintentionally lead to production delays, forced redesign or re-sourcing, failing compliance checks, profitability erosion, and even loss of market share.
Teams don’t need to trade velocity for resiliency because all the intelligence people need to make better decisions at the point of design already exist. The only question is how to make them available and in context of what they’re trying to do, at the point of design.
How does the pace of new product announcements change expectations on engineering, sourcing, and procurement teams?
Given the various success metrics that each team owns, new product announcements have a cascading effect on teams across the organisation.
For engineering, this means an increased focus on re-using parts, while at the same time, heightened pressure to increase the pace of innovation.
For sourcing and procurement, new product announcements highlight the need for these teams to ‘co-own’ the design of a product by influencing part selection early and helping identify optimal component choices, free of risks ranging from compliance to life cycle while achieving favourable cost positions.
It’s also important to develop strategic relationships with key suppliers. This allows for access to product roadmaps and production capacities when necessary.
As procurement teams seek to secure capacity, this creates additional pressure to collaborate with designers earlier in the cycle so they can de-risk and get a head start on locking in both pricing and supply commitments.
In this way, procurement needs to discover ways to de-risk the product launch and avoid creating bottlenecks due to lack of intelligence or insight.
Which component-related risks most commonly limit the commercial potential of AI-enabled products?
In this type of scenario, we see the promises of AI clash with the limitations of reality. Component-level restraints surrounding commodities like GPUs, high-bandwidth memory (HBM), and other high-performance components limit how quickly companies can scale production.
In this way, manufacturing AI-enabled hardware is a costly and high-risk pursuit in 2026, and will continue to be the case through 2027, based on expert predictions. This means that every AI-adjacent commodity will continue to face constraints around inventory and lead times, with near-constant pricing issues as well.
Beyond immediate issues, AI-related components also tend to have short or uncertain lifecycle timelines, which conflict with a market that expects long commercial lifetimes. Another concern is geopolitical issues like export controls on AI chips, tariffs that bring concerns surrounding country-of-origin to light, and limited potential markets.
What data points prove most valuable when sourcing intelligence is introduced earlier in the design process?
The key distinction here is not simply the type of data or the amount of it. Instead, what really moves the needle when it comes to sourcing intelligence is when insights are presented and in what context.
All the information in the world isn’t useful if it’s not relevant and tailored to the situation at hand. The right application of sourcing intelligence can transform a component decision from a guess into an informed and risk-aware decision.
The most important types of data points here include lifecycle status, compliance information, qualifications, supply availability, and pricing.
Another type of data point often overlooked in sourcing is form-fit-function alternates. Engineers will use these to quickly identify suitable components that match their requirements, but sourcing teams can also use this type of data to collaborate with engineering and mitigate risk as a result.
And of course, all of this focus on ‘earlier in the design process’ is part of a ‘shift left’ mindset in which teams focus on risk mitigation efforts earlier in the product lifecycle. This type of approach has gained momentum in recent years and will continue to gather support as time goes on.
How does real-time visibility of availability, pricing, and risk signals change engineer behaviour during component selection?
In short, it reframes and recontextualises the way engineers make component selections. The issue is providing properly contextualised insights based on real-time data that engineers can use in the moment.
Having access to all this data is great but we can’t expect engineers to know what to do with that information, because it’s not their job. It’s not that they don’t care but they don’t have time to ‘interpret’ those signals for their day to day.
This is where features like Supplyframe’s Risk Index do a great job of ‘packaging’ all of these signals into something as easy to understand as a traffic light, allowing engineers to quickly pivot to something less risky without spending hours or days trying to understanding supply chain risks.
Instead of focusing purely on supply constraints or vendor preference, this information allows engineers to effectively weigh performance and availability tradeoffs as well. It also means that they can actively avoid high-risk parts before they’re added to the BOM.
In addition to this, teams can also prioritise parts with multiple sources, and access design assets like footprints and 3D models to more effectively see how they fit into the larger board design.
All of this leads to smarter, more effective decisions from the beginning. It accelerates decision-making without sacrificing accuracy.
How does integrated sourcing intelligence help companies protect margins in the face of inflation and tariff uncertainty?
The presence of integrated sourcing intelligence means that things like inflation and tariffs are no longer surprise costs. Instead, they become design-phase variables, allowing teams to shape products and supply chains before volatility hits.
During the design phase, teams can use this intelligence to understand historical price volatility around specific components and often predict how that will manifest in the future. This enables them to design with cost elasticity in mind.
When it comes to challenges like tariffs, teams can identify the country-of-origin for parts and seek out alternate sourcing regions before parts are locked in.
Access to sourcing intelligence also means that alternate parts can be pre-qualified in the event that existing line items on the BOM need to be changed. This eliminates panic buying, which often comes at a significant price premium.
How do you see the relationship between AI-driven design and sourcing intelligence evolving over the next few years?
It’s likely that we’ll see AI-driven design and sourcing intelligence merge into a single platform in the coming years. The two will no longer co-exist, but rather, they will directly inform one another and enhance decision making as a result.
The idea that the design-to-source process is a linear and sequential action will be a thing of the past. Instead of one happening before another, they will happen simultaneously and in parallel, thanks to connected data and cross-functional collaboration.
As this intelligence layer informs every decision, AI models will continue to learn supply behaviour and best practices around specific design constraints and requirements. All of this will work in tandem to better contextualise and present relevant insights throughout the product lifecycle.
What capabilities should engineering and procurement teams prioritise to prevent this ‘digital divide’ from widening further?
To properly address the growing ‘digital divide’ between teams, the approach should be to think about how shared capabilities and evolved processes can holistically transform the entire product lifecycle.
Simply adding more processes isn’t going to solve the issue here. Instead, current processes need to evolve and transform to include real-time context and awareness around the factors that matter most.
Teams that transform the right decision into the easy decision will find success. Things like cost, risk, and lead times should be visible early and often across teams. This is the approach that will close the gaps and benefit the entire organisation.
