In the second episode of The Humans in the Loop, Tulip’s Pete Hartnett, Product Management Lead for AI, and Brennan Reamer, Ecosystem Engineer, join host Madilynn Castillo to explore a question at the center of modern manufacturing: how can AI make a real difference on the shop floor?
While AI often gets discussed in the abstract, this conversation centers on real problems in real factories — and the people who solve them. It’s about the practical ways AI can support operators and engineers: helping them access information faster, collaborate across languages, and reduce friction between people and systems, all while keeping human judgment at the core.
Much of Tulip’s progress in AI comes from this direct connection with users. Every product feature starts with feedback from the people who use Tulip daily — the engineers, operators, and supervisors who know exactly where traditional tools fall short. That insight gives Tulip’s approach to AI its strength: it’s not built in isolation, but alongside the people who understand the work.
Pete begins by acknowledging a truth familiar to anyone in industrial technology: AI can be both overhyped and transformative. The key, he explains, is to focus on the kinds of problems AI is uniquely suited to solve — those too complex, variable, or dynamic for traditional software approaches.
That mindset defines Tulip’s philosophy. The company isn’t chasing trends or building technology for its own sake; it’s solving for the specific conditions of manufacturing — where context, reliability, and safety matter as much as speed.
Manufacturing is this uniquely documented world. There are hundreds and thousands of SOPs and troubleshooting guides, but we don’t make that very readily accessible to the shop floor.
That observation, rooted in years of hands-on manufacturing experience, led to the creation of the Frontline Copilot — a chat-based interface that connects people directly to verified process knowledge. Trained only on a manufacturer’s own data, it allows operators to ask natural-language questions and instantly retrieve instructions, history, or troubleshooting steps.
Instead of another layer of complexity, it simplifies access to information — transforming static documentation into actionable guidance that helps people do their work better.
For Brennan Reamer, who leads demonstrations at Tulip’s Experience Center, the goal is to show how these tools come to life in real operations. The center highlights what happens when AI isn’t a separate initiative but part of everyday workflows — a natural extension of how teams communicate, learn, and respond to problems.
One of the most popular demos integrates the copilot widget into an Andon system. When an operator triggers an alert, they can open the chat, review past alerts, and learn how similar issues were resolved. “It retrains itself as you move on,” Brennan says. Each new incident contributes to a shared operational memory — a system that learns the same way people do.
Another demonstration addresses one of manufacturing’s most persistent barriers: communication across languages. “You can have a totally multilingual production line,” Brennan explains. “Tap an RFID badge, log in, and the app automatically changes from English to Spanish to French, all using AI translations.”
That simplicity underscores a deeper truth: effective AI design starts with empathy for the user. The best tools don’t replace people — they remove friction, letting expertise and creativity flow where they’re most needed.
As AI adoption accelerates, reliability becomes just as important as capability. In industries like life sciences, aerospace, and medical devices, trust is non-negotiable. Pete emphasizes that every AI feature at Tulip is designed with human oversight built in.
“When we’re working with a language model, we make sure there’s always a human in the loop to verify that output,” he says. “You can safely adopt this stuff without validation overhead or risk.”
That principle — human-in-the-loop by design — is central to Tulip’s philosophy. The company handles the technical complexity behind the scenes, from governance to validation, so manufacturers can experiment with confidence. Safety isn’t an afterthought; it’s what enables innovation to happen at scale.
Brennan recalls a recent visit from an automotive manufacturer still relying on printed work instructions. “They had hundreds of PDFs handed out to every operator,” he says. “Now they can use the AI Composer widget to build those into Tulip apps.”
In minutes, static files become interactive applications that capture data, trigger alerts, and connect to other systems like ERP or PLM. Engineers can then iterate — adding integrations, validations, and analytics — to continually improve those digital workflows.
Pete explains that this capability grew out of a common customer challenge. Many manufacturers begin their digital transformation by digitizing existing processes. AI Composer automates the repetitive part of that work, turning document-heavy workflows into flexible, data-driven applications.
By embedding these tools directly in Tulip’s platform, customers inherit version control, approvals, and security features automatically — critical for compliance-driven industries.
AI’s promise lies not in one-off efficiency gains but in continuous improvement. Pete describes Tulip’s development process as iterative by design: start with a hypothesis, test it in real environments, and learn from every deployment.
That feedback loop mirrors how customers use the platform. Teams experiment, refine, and scale — evolving their operations one workflow at a time. It’s a cultural shift as much as a technical one: innovation becomes part of standard work.
One example came from customer feedback on AI Composer. Long-time users asked for the ability to apply Composer to existing app templates, merging automation with their own best practices. Within months, Tulip built the feature — a direct reflection of how human insight continues to guide the platform’s evolution.
Not every problem requires a large-language model. Some of the most valuable capabilities are also the simplest. “Some of these atomic capabilities — like translation or speech-to-text — are incredibly valuable, but they’re not flashy,” Pete says. “They’re really solving problems.”
That pragmatism defines Tulip’s approach to AI. The goal is fit, not flash: choosing the right technology for each challenge and designing with the end user in mind. The best outcomes come not from automation alone, but from systems that make complex work more intuitive.
As Brennan observes, visitors to the Experience Center often leave inspired not by futuristic visions of robotics, but by the elegance of everyday solutions — tools that close communication gaps, simplify troubleshooting, or make collaboration effortless.
This episode of The Humans in the Loop captures the essence of Tulip’s mission: building technology that amplifies people rather than replaces them. “It’s really about how do you give that human superpowers,” Pete says, “help them do the thing they’re uniquely good at, and give them leverage on the things that don’t require their full skill set.”
That balance — humans leading, AI assisting — defines Tulip’s vision for the future of manufacturing. From copilots that surface knowledge in seconds to translations that connect global teams, the most effective AI isn’t abstract or autonomous. It’s the kind that works quietly alongside people, embedded in the rhythm of daily work.
Q: How is Tulip applying AI on the shop floor?
A: Tulip embeds AI directly into workflows through tools like Frontline Copilot, AI Composer, and AI Translations. These features help operators access information, digitize documentation, and collaborate seamlessly across languages — all within a governed, human-in-the-loop framework.
Q: Why is human-in-the-loop AI important for manufacturing?
A: It ensures that AI remains trustworthy and auditable. Humans validate outputs, preserving compliance and safety while benefiting from automation.
Q: What industries benefit most from Tulip’s AI tools?
A: Manufacturers in regulated sectors — such as medical devices, aerospace, and pharmaceuticals — as well as automotive and consumer goods producers seeking faster digitization and safer AI adoption.
👉 Learn more about Tulip AI and explore how you can bring practical, human-centered AI to your operations: Tulip AI Product Page
