Mija Tortilla Factory’s AI Tools are a Key Selling Point for Retail Buyers  11/19/2026


With Enzo Lihard

Enzo Lihard, VP of Sales and Director of Operations for Mija Tortilla Factory, has a unique pitch for retail buyers he meets with at ECRM Sessions. Sure, the company makes great tortillas, and produces them on a massive scale. But what really grabs the buyers attention is the fact that Lihard – a software engineer with nearly 30 years in the pharmaceutical and biotech industries – built a proprietary, multi-agent artificial intelligence platform called Tracecore Food that monitors, validates, and documents every step of production in real time.

In other words, he can tell them exactly when the flour is being added to a batch that will soon end up on that retailer’s shelves, and knows that every ingredient has been tracked, traced and is completely compliant with all food safety standards. 

What’s more, he is making the Tracecore Food platform available to other brands, so that the industry as a whole can benefit from enhanced supply chain visibility. I caught up with Lihard at our recent ECRM Food & Beverage Session in Dallas to learn more about Mija Tortilla Factory and the innovative platform he built. You can watch our full video interview below!

From Big Pharma to the Tortilla Line

Before worrying about dough hydration, masa pH levels, or oven thermocouples, Lihard spent almost three decades developing mission-critical software for major life sciences companies.

"Basically, I'm a software engineer," Lihard explains. "I used to work for the life science industry for over 28 years – big players like Eli Lilly, Johnson & Johnson, Baxter, Amgen. And I also traveled quite a bit all over Europe. I was involved with GlaxoSmithKline implementations in England, as well as in Germany, Spain, and Italy, before coming back to the States."

Working across global regulatory environments (he speaks four languages) instilled an uncompromising standard around process validation and data integrity. Eight years ago, that background intersected with the tortilla business when Mija Tortilla Factory’s founders ran into trouble with programmable logic controllers (PLCs) and legacy machinery.

"Eight years ago, they had some problems with PLCs and some legacy equipment," Lihard recalls. "So I jumped in to help out and pinpoint some of the issues. From there, we built a great relationship, and a year later I developed my first piece of software for them, which was an SPC – Statistical Process Control – system. I have always loved numbers, and I have always loved working with data. Data is the most valuable thing that we have."

Today, the Louisville facility can produce over 67 million corn and flour tortillas a month – from six-inch street-taco sizes up to 14-inch burrito wraps – alongside tortilla chips. Mija Tortilla Factory supplies foodservice operators and retail grocers, including its retail label, La Sabrosita, offering white and yellow corn tortillas with traditional recipes and blue corn varieties coming soon, while holding Kosher and SQF Level 3 certifications.

"The SQF certification is crucial for every single manufacturer today," Lihard said. "If you don't have that, you're out of the game. Earning and maintaining SQF allowed me to learn deeply about food regulation, and my background helped me enormously because life sciences such as pharma, biotech, and medical devices represent the highest-regulated industries in the world. If we do something wrong there, people can die. So I brought all of that discipline and rigor directly into my food operation."

Bridging Data Silos With Tracecore Food AI Agents

At Mija Tortilla Factory, production spans flour tortillas, corn tortillas, and chips, supported by optical vision systems, inline checkweighers, end-of-line metal detectors, lab pH readings, and 800-degree ovens. Traditionally, food plants store this data in isolated spreadsheets, machine logs, and paper binders. Lihard saw early on that disconnected databases could not keep pace with high-speed production.

"Traceability is one of the biggest priorities for every single manufacturer," Lihard said. "It's not just about when you make a batch of tortillas, it's also about knowing your supplier's source. Are you validating that? Do you validate your process? Are you constantly checking your metal detector? Are you constantly verifying the quality certificate for your flour and for every ingredient you use in your operation? I started tracking every single step of my process, from the vision system and the weight system back to my lab with pH readings and all the dynamic parameters we have to monitor. One of my colleagues always told me, 'Enzo, it is a fine dance with all of these processes, and you have to stay in step.' So I said to myself, 'Instead of keeping every single database separate, I want to build something that builds bridges.'"

That vision became TraceCore Food, a 100 percent cloud-based ecosystem connecting receiving, engineering, quality assurance, maintenance, and shipping – allowing Lihard to monitor live plant activity even from his meeting space at the ECRM Session in Dallas.

At ECRM's Food & Beverage Session

"One of the things that is very important to me is being able to be away for business while still staying completely connected inside the facility," Lihard says. "The software is 100 percent portable; everything is in the cloud. For example, this morning back in Louisville, we are in the middle of a deep cleaning on one of the lines, and I just received an alert from one of my agents telling me the deep cleaning has started so I can get ready to review those quality assessments. My dashboard tells me, 'Here is your daily assessment, here is your weekly assessment, here is your sanitation assessment, and here is your environmental monitoring.' Everything is connected."

Inside the Multi-Agent AI Ecosystem

Rather than relying on a single algorithm, Lihard designed TraceCore Food around specialized AI agents trained on verified primary sources.

"To become an expert, an agent needs to know the exact rules," Lihard says. "First, you have to train your agent on the right foundation. For the SQF Institute standards, I downloaded the entire Edition 9 code, and now Edition 10, and built an agent fully trained on that primary source. For anyone who wants to start building with AI, you have to feed reliable, authoritative sources into your agent, because it really is garbage in, garbage out. Just like in food, we are what we eat."

With a regulatory agent anchored in official SQF documentation, Lihard runs cross-reference analyses comparing SQF Edition 9 against Edition 10 to pinpoint updated KPIs, risk evaluations, and vulnerable process steps. Alongside that compliance intelligence, operational agents ingest data from PLCs, lab tests, and floor terminals:

  • Order and Production Scheduling Agents: Align customer orders with runs across flour, corn, and chip lines.
  • Preventive Maintenance Agents: Monitor machine logic – from pumps and heat plates to oven thermocouples – to schedule maintenance and prevent downtime.
  • Food Safety and Metal Detection Agents: Verify mandatory end-of-line metal detector checks for every SKU, compiling reports every six to eight hours.
  • Receiving, Shipping, and Logistics Agents: Track pallet inspections, supplier flour certificates, bills of lading, and finished-goods stock.

Turning Frontline Operators into ‘Live Agents’

When rolling out TraceCore Food in Louisville, Lihard faced two realities: personal cell phones are banned on the plant floor for food safety reasons, and most operators speak Spanish. Rather than forcing workers onto English-heavy forms, Lihard equipped floor staff with tablets running a visual, color-coded messaging interface modeled on everyday mobile apps.

"In my factory, cell phones are not allowed on the floor, so I built an internal messaging system where every single operator has access to a tablet," Lihard says. "For the operators, making it intuitive was the biggest challenge, because while I have a few team members who are bilingual, most of them speak Spanish. So I designed the application around the way we use our phones today: very friendly to use, with clear buttons, color coding, and open capabilities to snap photos and record notes. In a very real way, my operators become like live agents inside the process."

When an operator spots an issue, such as a damaged ingredient super sack, they snap a photo and log a note in Spanish.

"An operator can see something that doesn't look right, snap a picture immediately, and message me directly: 'Enzo, do you think this is right?'" Lihard said. "I can look at the photo from anywhere, see the visual evidence right away, and either walk out to the line myself or dispatch a technician immediately."

To eliminate language friction, Lihard built a translation AI agent into the workflow. "I sit down with my team, have lunch and dinner with them, and tell them, 'We want this environment to be easy and friendly for you, but the most important thing is that you document everything,'" he says. "'I don't care which language you write in. Put whatever you observe into the tablet, and I have another agent that automatically translates every message into plain English so I can integrate that data straight into my master batch report.' I always tell my team, 'Every single detail matters, guys, because it's a story. If you lose a detail, the story has something missing. Our batch record is the complete story of how we made that tortilla.'"

Pharma-Grade Traceability and Leaner Operations

Under SQF rules and federal regulations, if a supplier reports an issue with a preservative or flour lot, Mija Tortilla Factory must trace and isolate every affected case in under 24 hours.

"If one of my suppliers calls and says, 'Enzo, we put something wrong in your preservative,' I need to be able to pull that entire trace in less than 24 hours," Lihard explained. "Without automation, how do you do that? You need a large team—six to eight people minimum—just digging through binders to pull all those records together. So I brought the pharmaceutical batch-record model into my operation. I created a comprehensive electronic batch record that captures not only our exact formulation, but every specific ingredient lot we used, linked directly back to the supplier documentation."

Because TraceCore Food tracks every case from raw material receipt through shipment, Mija Tortilla Factory can isolate a single affected batch instead of discarding healthy inventory.
The labor savings across five or six high-speed lines are equally compelling. According to Lihard, manual tracking requires tons of paper, endless forms and at least four to six QA technicians just walking the floor to complete hourly or bi-hourly inspections required by SOPs. 

“Meanwhile, if a line supervisor is dealing with a jam on a high-speed machine, the last thing you want is for them to juggle a mechanical jam plus a clipboard plus walking back to sign forms,” says Lihard. “TraceCore Food easily saves between 10 and 15 full-time people in administrative and redundant tracking overhead. I can run efficiently with one supervisor per shift, one QA technician per shift, and my senior engineers focused on supporting the lines and preventing downtime."

Lihard also structured the platform around pharmaceutical ALCOA data integrity principles, ensuring records remain Attributable, Legible, Contemporaneous, Original, and Accurate from day one.

"We protect the original source data from day one, and that is critical for any audit, whether it's an FDA audit or an SQF audit," Lihard noted. "Auditors love how transparent it is. I can literally hand a single tablet to an auditor and say, 'Everything is right here. Let me know exactly what you want to review, and let's walk through it.'"

Radical Transparency as a Competitive Moat

For mid-sized food manufacturers pitching chains like Circle K or H-E-B, overcoming buyer skepticism is essential. But Lihard tells buyers that he doesn’t just make tortillas, he makes data, and the outcome of that data is a fantastic and very safe product. 

To back up that claim, Lihard gives high-volume partners access to a real-time SCADA and Statistical Process Control portal via a secure browser link.

"For our top, high-volume customers, we built a real-time data monitoring portal," Lihard says. "You can see your SKU in real time – how we are producing it, the mixing parameters, the oven temperatures, the pH of the tortilla, the weights, and all the quality checks along the batch, alongside Statistical Process Control metrics showing how effective our lines are. It's all about confidence and perception. When a buyer is considering giving you a multi-million-dollar purchase order to supply 14,000 stores across the country, they aren't going to hand you that business just because you have a pretty label. They need to trust your operation. When I show buyers that live dashboard, their reaction is, 'Wow, what is this?' It proves we are a reliable manufacturer where they can place massive volume and know the process is always under control."

What’s Next: Opening TraceCore Food to the Industry

Following enthusiastic feedback at ECRM, Lihard confirmed he is preparing to commercialize TraceCore Food for the wider food and beverage industry. "Right now, my next step is continuing serious conversations I have already started with investors," Lihard says. "We want to make this software available to everyone in the industry. Based on everything I heard here, I have an important meeting with my team to launch a free AI agent directly on TraceCoreFood.com."

That upcoming public tool will be a free compliance agent trained on SQF Edition 10, helping food safety managers and plant operators ask questions, compare code updates, and map out certification steps at no cost.

"It will be a free expert agent on SQF 10," Lihard says. "When manufacturers start asking questions, comparing requirements, and figuring out how to achieve that level of certification, having an expert tool like that is hugely beneficial, and it's something I would love to share and give for free to every single person in the industry."

From troubleshooting legacy PLCs in Louisville to architecting an AI-driven traceability platform built for 14,000-store retail rollouts, Lihard’s journey offers a clear blueprint for modern CPG manufacturing. Great recipes win over consumers at the table, but pairing culinary quality with radical, AI-powered data transparency is what earns the trust of the world’s largest buyers.

 

 

Joseph Tarnowski

VP Content
ECRM

Post a Comment

Log in to Comment