91������

Core problems for industrial manufacturing companies are not new: How to lower attrition rates? How to improve customer satisfaction? How to increase productivity? How to increase energy efficiency? ��

‍

What is new, though, are a set of early-stage companies springing up to tackle these problems using technologies that weren’t available even a few years ago. They are approaching these problems with surprising and shockingly effective results. ��

‍

In Ayna’s new podcast series we talked with a handful of founders at AI companies. This article provides insights into these conversations and will summarize key themes to think about as these companies take the stage.

‍

Using new AI to solve old problems

‍

Customer service and outbound communication. Adam Earle is the CTO of Tenyx, and he’s working to create AI voice solutions to transform the way industrial technology companies handle customer service and outbound communication. Earle views outdated interactive voice response systems as outdated, so he focuses on providing practical advice to companies who are working to use LLM’s to create natural and intelligent conversations. This way they won’t sound like robots. ��

‍

Customer service and other outbound communications are specifically ripe for disruption because of its short-term ability to free up humans to focus on more complex enterprise issues. This isn’t something that is still a few years away like so many topics and challenges that many AI companies are working on—this is ready today. ��

‍

Listen to the full conversation with Earle .

‍

Talent Acquisition. As the CEO of Eightfold, Ashutosh Garg understands how AI will affect talent management and workforce development. Specifically for industrial manufacturing, where there have been high attrition rates, Eightfold helps companies build a network of talent, “their own talent pool,” to create seamless and efficient internal development. If a company gets shut down, Eightfold helps build a network of career supporters who can help upskill the unemployed person. ��

‍

Eightfold uses data to better understand what people have done in their current jobs. Then they use that data to understand what the employee can do next. It’s a data problem that they’re working to fix, but Garg says that the most fundamental problem is the need for talent, and that hasn’t changed:

‍

“In the last five years, we’ve seen the highest unemployment rate ever, and we had the lowest unemployment rate ever…those market analytics will keep on changing. But what has not changed is people need jobs. And companies need people to grow and succeed.”

‍

Listen to the whole episode with Garg .

‍

Manufacturing Productivity. We spoke with John Herlocker at Tignis, a startup focusing on AI driven solutions for industrial automation. Herlocker emphasizes the importance of having a defined market and buyer before product development, and over a seven-year period with Tegnas, they’ve shifted from a broad focus to specializing in semiconductors. ��

‍

Semiconductor manufacturing faces many unique complexities. High costs and even the slightest variability in construction processes can leave a significant number of chips on a wafer non-functional. So Tegnas looked to AI to manage these variabilities, reduce time to market, and lower costs. They’ve partnered with equipment manufacturers to offer them technology that allows for better control of the processes.

‍

Herlocker remains cautious, though, about incorporating generative AI into the unusually conservative sector, and they know the importance of precision in any advice provided by AI. Herlacker’s overarching goal at Tegnas is to make sure that as the semiconductor market grows, the manufacturing processes can keep pace. ��

‍

Listen to our conversation with Herlocker

‍

Automation. Antti Karjalainen founded Sema4.ai to leverage generative AI to help companies build their own AI Agents. Initial use cases surround internal operations and customer support, where manual work can more easily be turned over to AI. Karjalainen points out that industrial manufacturing companies are adopting Agents quickly. This is due to favorable regulations compared to industries like banking and insurance. ��

‍

Most initial adoptions in industrial manufacturing companies are being used for paying invoices to vendors that internal teams are currently doing manually. To reduce this manual workflow, Karjalainen is using AI to automate tasks like document processing and other financial processes. Sema4 is also helping companies program Agents for more complicated compliance work in finance, and finally, using their AI to identify and invent new revenue streams.

‍

Karjalainen says keeping his products open source is fundamentally important to him as a founder of an upstart AI company:

‍

“The speed of innovation is truly tremendous and building on an open-source foundation allows us to tap into that breathtaking speed of innovation in AI technologies.”

‍

Listen to the full conversation .

‍

‍

Training and Development. Alex Ratner is the CEO of Snorkel AI, a company that transforms AI development by labeling and curating data to use it to teach, adapt and evaluate AI models. Ratner says the data is often the messiest and most critical part of AI, and for Snorkel, the most important challenge to solve.

‍

Ratner explains in our conversation that the biggest use cases for AI are around extracting certain things from unstructured data like text or image data, and then tagging, classifying, and summarizing that data. ��

‍

If you want to use AI to do all this parsing and leveraging of your unique data, you need to teach it to work well with your data. That’s where Snorkel comes in. Since data is the key interface for getting AI to learn anything, Snorkel’s mission is to make that data labeling for AI programmatic, the same way it is for software development. Their goal is to make it “fast, iterative, auditable, and adaptable, instead of just one click at a time”, the way it is today for most companies. They want a high degree of automation, but Ratner adds, “still with a human in the loop to codify their knowledge.”

‍

Listen to our conversation with Ratner .

‍

Implications and Effects on Industrials

‍

These AI companies are just getting started, but they already present a significant opportunity for industrial manufacturing companies to embrace the innovations and opportunities they are creating. Otherwise, tackling traditional challenges and improving operational processes will lag behind your competitors.

‍

It’s worth considering, though, that successful integration of these technologies will hinge on your company’s willingness to adapt and reimagine your existing processes and workflows. ��