[ad_1]
AI agents have become very popular, a trend driven by the boom in generative AI and large language modeling (LLM) in the past few years. Getting people to agree on what AI agents are is a challenge, but most maintain that they are programs that can be given tasks and given decisions to make – with varying degrees of autonomy.
In short, AI agents go beyond what a chatbot can do: they help people get things done.
It’s still early days, but the likes of Salesforce and Google are already investing heavily in AI agents. Amazon CEO Andy Jassy recently hinted at a more “agent” Alexa in the future, which is as much about action as it is about words.
In parallel, startups are also raising money from the hype. The last of these companies is the German company Juna.aiwhich wants to help factories be more efficient by automating complex industrial processes to “maximize production productivity, increase energy efficiency, and reduce overall emissions.”
To achieve this, the Berlin-based startup said today that it has raised $7.5 million in a seed round from a Silicon Valley venture capital firm. Kleiner PerkinsBased in Sweden Norskin VCand Chairman of Kleiner Perkins John Doerr.
Self-learning is the way
Founded in 2023, Juna.ai is hand-made Matthias Auf der Maur (Pictured above, left) and Christian Hardenberg (Pictured above, right). Der Mauer previously founded a predictive machine maintenance startup called AiSight and It was sold to the Swiss smart sensor company Sensirion In 2021, while Hardernberg was the former CTO of European food delivery giant Delivery Hero.
At its core, Juna.ai wants to help manufacturing facilities shift to smarter, self-learning systems that can achieve better margins and, ultimately, a lower carbon footprint. The company focuses on so-called “heavy industries”, which are industries such as steel, cement, paper, chemicals, wood and textiles with large-scale production processes that consume a lot of raw materials.
“We work with process-driven industries, which often involve energy-intensive use cases,” Der Maur told TechCrunch. “So, for example, chemical reactors that use a lot of heat to produce something.”
Juna.ai integrates with manufacturers’ production tools, such as industrial software from Aviva or Suband views all its historical data obtained from the machine’s sensors. This may include temperature, pressure, speed, and all specific output measurements, such as quality, thickness, and color.
Using this information, Juna.ai helps companies train their in-house agents to know the ideal machine settings, providing operators with real-time data and guidance to ensure everything is running at peak efficiency with minimal waste.
For example, a chemical plant that produces a special type of carbon might use a reactor to mix different oils together and subject them to an energy-intensive combustion process. To maximize production and minimize residual waste, conditions must be ideal, including the levels of gases and oils used, and the temperature applied to the process. Using historical data to determine optimal settings and taking real-time conditions into account, Juna.ai agents are supposed to tell the operator what changes they should make to achieve the best results.
If Juna.ai can help companies fine-tune their production equipment, they can improve their productivity while reducing energy consumption. It’s a win-win, both for the customer’s bottom line and their carbon footprint.
Juna.ai says it has built its own custom AI models, using open source tools such as TensorFlow and By Torch. To train its models, Juna.ai uses reinforcement learning, a subset of machine learning (ML) that involves a model learning through its interactions with its environment — it tries different actions, observes what happens, and improves.
“The interesting thing about reinforcement learning is that it’s something that can take action,” Hardenberg told TechCrunch. “Models only make predictions, or maybe they generate something. But they can’t control.”
Much of what Juna.ai does nowadays is akin to a “co-pilot” – it presents a screen that tells the operator what adjustments they should make to the controls. However, many industrial processes are incredibly repetitive, which is why enabling the system to take actual action is beneficial. For example, the cooling system may require constant fine-tuning to ensure the device maintains the proper temperature.
Factories are well accustomed to automating system controls using Product ID and Monetary Policy Committee Consoles, this is something Juna.ai can do as well. However, for an AI startup, it’s easier to sell a co-pilot, it’s baby steps for now.
“It is technically possible for us to allow it to operate independently for now; we would just need to implement the connection. But in the end, it’s all about building trust with the customer,” Der Maur said.
The benefit of the startup’s platform is not labor savings, Hardenberg added, noting that factories are already “efficient” in terms of automating manual processes. It’s all about improving those processes to reduce costly waste.
“There’s not a lot of gain by removing one person, compared to a process that costs you $20 million in energy,” he said. “So the real win is: Can we go from $20 million in energy to $18 million or $17 million?”
Pre-trained agents
Right now, Juna.ai’s big promise is an AI agent tailored to each customer using their historical data. But in the future, the company plans to offer off-the-shelf agents that are “pre-trained” and don’t need as much training on new customer data.
“If we build simulations over and over again, we’ll get to a place where we can have reusable simulation templates,” Der Maur said.
So, if two companies use the same type of chemical reactor, for example, it may be possible to leverage and transfer AI agents between customers. One model of one device is the general essence.
However, one cannot ignore the fact that companies have been reluctant to dive into the burgeoning AI revolution due to data privacy concerns. Those concerns have been lost on Juna.ai, but Hardenberg said it hasn’t been a big issue so far, partly because of its data location controls, and partly because of the promise it offers to clients of unlocking the latent value from huge banks of data.
“I saw that as a potential problem, but so far it hasn’t been a big problem because we leave all the data in Germany for our German customers,” Hardenberg said. “They set up their own server, and we have top-notch security safeguards. On their side, they have all this data out there, but they haven’t been very effective at creating value from it; it’s been mostly used for alerting, or maybe some manual analytics. But our point is That we can do more with this data, like build a smart factory, and become the mastermind of that factory based on the data they have.
Just over a year after its founding, Juna.ai already has a handful of clients, though der Mauer said he’s not at liberty to reveal any specific names just yet. However, they are all based in Germany, and they all have branches elsewhere, or are subsidiaries of companies based elsewhere.
“We plan to grow with them, it’s a very good way to expand with your customers,” Hardenberg added.
With a new $7.5 million in the bank, Juna.ai is now well-funded to expand beyond its current headcount of six, with plans to double its technical expertise.
“It’s a software company at the end of the day, and that basically means people,” Hardenberg said.
[ad_2]