Tellspec - Beam Your Health Up!

In the world of data-driven solutions and cutting-edge technology, few have managed to bridge the gap between complex science and real-world applications as effectively as Isabel Hoffmann. As the Founder and CEO of Tellspec, Isabel has revolutionized the way we approach organic substance analysis, utilizing biophotonic sensors, machine learning, and predictive software models. Tellspec is not just a company; it’s a partner in industries ranging from food safety and pharmaceuticals to archaeology and veterinary science. With over 11 years of expertise in spectroscopic technology and machine learning, the company has developed a robust portfolio of tools designed to test the quality, authenticity, and even the decay of organic substances. In this exclusive interview with Isabel Hoffmann, Delivery Rank delves into the cutting-edge solutions Tellspec offers, her journey as a tech entrepreneur, and the exciting future of biophotonics and machine learning in the world of organic substance analysis.

You’ve been at the forefront of merging spectroscopy with machine learning. What initially sparked your interest in this intersection of science and technology, and how did that evolve into Tellspec?

My interest in the intersection of science and technology came from the merging of my academic career—I am a mathematician and was for eleven years the an associate professor and director of a large information technology centre at the University of Toronto —and my experience as an entrepreneur, having founded and scaled eight companies over nearly three decades. Tellspec, emerged from a more personal concern: the critical need for food safety and transparency. 

As a parent I was troubled by the lack of accessible tools to understand what was in the food we consume daily. The idea that advanced scientific techniques could be made a food testing device portable and user-friendly fascinated me. Spectroscopy, with its ability to reveal molecular fingerprints, seemed like the perfect tool, if only it could decode complex data more efficiently and be made available to consumers in real time. This challenge is where machine learning came in. By training algorithms to interpret spectral patterns, we transformed sophisticated laboratory techniques into portable, on-the-field, and practical applications. Early experiments proved that even smartphones could eventually become spectrometers. This vision evolved into Tellspec: a platform designed to empower individuals and businesses to analyze food composition instantly, testing food for quality, authenticity, and safety.

What began as a mission to protect my family evolved into a grander global initiative: to uncover the truth, rebuild confidence in the food supply chain, and empower consumers with knowledge. Tellspec isn’t just a tech-company, it embodies a movement committed to transparency, traceability, and informed decision-making in what we consume.

Tellspec has developed multiple sensors tailored to different needs—from preemies to enterprise use. Can you walk us through how user feedback influenced the design and evolution of these devices?

At Tellspec, user feedback not only guides our technology, it builds it. Each sensor we’ve developed has been shaped by actively listening to the individuals who rely on them. Whether combating food fraud, enhancing quality control, or improving neonatal nutrition, our sensors have evolved through direct collaboration with the users who depend on them most.

Here’s some examples of our critical areas:

“Food Quality: From Farm to Fork”

Feedback from agronomists and retailers revealed a glaring problem: that “premium” produce was judged subjectively based on visual inspection alone. We worked with several large retailers in Europe to train our algorithms on hard metrics—Brix, dry matter content and titratable acidity—we turned “premium” quality and shelf-life into an objective standard. Now, negotiations run on data, not guesswork. 

“Fighting Illicit Trade”

When we first engaged with a large tobacco company, they didn’t just want to detect adulteration—they needed to trace these counterfeits back to the source. Their team needed a tool which could distinguish authentic tobacco from counterfeits at the port of entry, not weeks later in a lab. This pushed us to refine our spectral libraries with region-specific markers and integrate blockchain for tamper-proof supply chain logging. Now, a single scan can flag suspicious deviations and link them to shipment records without lab delays.

“Preemie Nutrition: Precision for the Most Vulnerable”

Professionals working in neonatal clinicals told us, “We don’t just need macronutrient data—we need to see specific fatty acids and milk osmolality”. Their input enabled us to develop the only existing technology currently meeting the 2022 ESPGHAN  (European Society for Paediatric Gastroenterology, Hepatology and Nutrition) guidelines for human milk fortification, with capabilities to analyze multiple fatty acids and measure osmolality. 

The common thread through all these examples: actionable insights in real time. Collaboration helps us innovate. We have built customizable dashboards and API integrations tailored to each user’s workflow. 

More recently, Tellspec has been developing and deploying predictive analytics. Imagine a sensor combined with AI-technology to alert a food importer to rising fraud patterns in a specific region, or a NICU device that suggests nutrient adjustments based on a preemie’s metabolic response or the disease risk scores the infant may have.

Your platform claims to simplify laboratory analysis of organic substances—how do you strike the balance between precision and ease-of-use, especially for non-scientific users?

At Tellspec, we believe that precision should never come at the cost of accessibility. Our platform is designed to bridge the gap between laboratory-grade accuracy and the simplicity of a consumer device. Here is how we achieve that balance:

Spectroscopy is inherently complex, but users shouldn’t need a PhD to interpret it. We’ve done this with three smart layers: 1. Simple Hardware: Our portable sensors capture high-resolution spectral data—comparable to benchtop lab equipment—but with one-button operation 2. AI-Driven Translation: Our machine learning does the heavy lifting, converting spectral fingerprints into plain-language insights (e.g., "This olive oil contains 12% adulterated soybean oil"). 3. Contextual Guidance: Instead of raw numbers, users get actionable feedback or compliance alerts tailored to their industry.

We design for real-world needs, not just laboratories. Farmers and food inspectors testing for fish freshness don't need to see complex spectral data, instead they receive instant plain-language results such as: "the fish is safe to eat raw" , or  "cook thoroughly", or “discard, the fish is spoiled". Meanwhile, custom officials and food fraud investigators can use "Fraud Detection" modes, which auto-flag adulterations/anomalies without manual analysis, and neonatal professionals in NICUs receive straightforward fortification instructions (ex. fortification needed: +04g protein per 100mL) no biochemistry expertise required.  Every solution is designed to turn advanced science into immediate, actionable decisions.

We offer laboratory precision with pocket-sized ease. Our Calibration-as-a-Service automatically updates sensor calibrations based on global data trends (ex. new adulteration methods in the olive oil trade), so users always have cutting-edge detection without manual tweaks. All results are cross-validated against multiple reference databases, minimizing false positives. A coffee trader might see: "85% Arabica (confidence: 93%)"—transparent enough for experts, clear enough for buyers. A farmer can scan a grain shipment in seconds and reject it before unloading. A neonatologist can verify donor milk nutrients and receive a suggested targeted fortification in less than 5 minutes.

Tellspec’s technology enables real-time decision-making. What are some surprising or unexpected ways clients have used your tools that you hadn’t initially envisioned?

One of the most thrilling aspects of deploying spectral technology is watching users redefine its possibilities—often in ways we never imagined. While we built Tellspec for food safety and healthcare, our clients’ creativity has stretched its applications far beyond our original vision. Here are two of the most surprising use cases that left even our team in awe:

4.1. Archaeology Meets AI: Dating Ancient Bones in Caves

A university research team approached us with a radical idea: could our handheld spectrometers detect collagen degradation in fossils to estimate their age? Traditionally, radiocarbon dating requires destroying samples and waiting weeks for lab results. But by training our algorithms on spectral signatures of collagen residues—which break down predictably over millennia—they developed a non-destructive, real-time field tool. Now, archaeologists scan bones deep in caves and get instant estimates of their age range, revolutionizing how they prioritize excavations. We never dreamed our food-quality tech would help unravel human history!

4.2. Fighting Illicit Trade in Contraceptives Across Africa

Another client in West Africa repurposed our devices to combat counterfeit contraceptives flooding local markets. Their team used Tellspec’s sensors to scan glass bottles containing progesterone. They also integrated our API with a crowdsourced map app, letting pharmacies report counterfeit batches in real time. What began as a food fraud tool is now protecting reproductive health—proof that spectral transparency can save lives in unexpected ways.

The Common Thread? Spectral Data as a Universal Language.

These cases taught us that every organic substance tells a story—whether it’s a 5,000-year-old femur or a contraceptive. Now, we actively encourage this ‘hacking’ of our tech: Wine counterfeiters in France use it to verify vintage years by analyzing tannin oxidation, or textile recyclers sort organic vs. synthetic fabrics at scale using spectral ‘fingerprints.’

One should never underestimate human ingenuity—especially when you put powerful tools in the hands of those solving urgent, real-world problems.

Looking ahead, where do you see the biggest opportunities for Tellspec’s technology to make a societal impact—particularly in underserved or high-risk communities?

We designed our Preemie System to address the critical needs of underserved regions bearing the highest global preterm birth rates –where NICUs often lack even basic diagnostic tools. By democratizing access to lab-grade nutritional analysis we are not just improving care, we are transforming survival outcomes for the most vulnerable newborns. 

With 15 million preterm births annually (60% in Sub-Saharan Africa and South Asia), clinicians in low-resource settings face an impossible choice: fortify breast milk blindly—risking deadly underfeeding (necrotizing enterocolitis) or overfeeding (metabolic strain)—or skip fortification altogether. Additionally donor milk screening remains virtually nonexistent. . Our handheld sensor addresses this by delivering lab-grade nutritional analysis (calories, protein, fat, fatty acids) in 20 seconds at <70% the cost of traditional systems—requiring no infrastructure, battery-powered offline-capable and operable by nurses with minimal training. Beyond individual care, anonymized aggregate data can reveal regional malnutrition patterns (e.g., protein deficits in drought zones) to guide public health interventions. 

We are now partnering with NGOs and local manufacturers to localize production of sensors in economically developed countries so we can cut costs. Seamless telehealth integration would enable automatic upload of scan data to cloud platforms, allowing remote neonatologists to review and adjust fortification plans in real-time. Currently we are working on our predictive capabilities as our research suggests the Preemie Sensor could analyze maternal urine samples to identify preeclampsia risk factors, potentially enabling life-saving early interventions. Why does this matter? In high-income countries 80 to 90% of preemies survive, while in low-resource settings, as low as 10% will survive. There is a huge disparity here. Tellspec’s job is to transform invisible nutritional truths into actionable insights, for every preemie, everywhere. We are helping build a world without health inequity for preemies.

To read more about Tellspec, please visit https://tellspec.com/