A significant stride has been made in the quest for more palatable and sustainable protein sources with the recent announcement that NECTAR, the non-profit arm of the philanthropic investment platform Food System Innovations (FSI), has secured a $2 million grant from the Bezos Earth Fund. This substantial funding will accelerate NECTAR’s ambitious project in collaboration with Stanford University to develop a sophisticated AI model. This AI aims to bridge the gap between molecular structure, sensory experience (flavor and texture), and consumer preference, with the ultimate goal of empowering the food industry to create alternative proteins that are not only better for the planet but also more appealing to consumers, thereby driving market adoption.
The initiative addresses a critical bottleneck in the burgeoning alternative protein sector: taste and texture. While consumer demand for sustainable food options is on the rise, widespread adoption has been hampered by the perception that many plant-based and other alternative proteins fall short in replicating the sensory qualities of their animal-based counterparts. NECTAR’s approach, grounded in extensive consumer sensory data, seeks to demystify the complex relationship between ingredient composition and consumer satisfaction, promising to streamline the product development process for food companies.
The Genesis of an AI-Powered Food Revolution
NECTAR’s journey toward this AI-driven solution began over eighteen months ago. Recognizing the immense potential of the unique sensory response data they were meticulously collecting, the organization initiated discussions with computer science experts at Stanford University. The objective was to explore innovative ways to leverage this novel dataset. This exploration coincided with the opening of the first phase of the Bezos Earth Fund’s AI Grand Challenge, a program dedicated to fostering technological solutions for climate and nature preservation.
Adam Yee, chief food scientist involved in the project, facilitated initial connections between NECTAR and the Stanford team. Their early work focused on the application of Large Language Models (LLMs) to support food scientists. This foundational research culminated in a paper presented at the International Conference on Machine Learning (ICML) in January, the world’s premier machine learning conference. The success of this initial phase paved the way for NECTAR to apply for and ultimately receive the significant Phase Two grant from the Bezos Earth Fund, announced in October.
Caroline Cotto, Managing Director of NECTAR, elaborated on the project’s core mission during a recent interview. "One of the things Nectar is doing is we just won a $2 million grant from the Bezos Earth Fund to take our sensory data and build a foundation model that will predict sensory," Cotto explained. "So we kind of bypass the need for doing these very expensive consumer panels, and then also predict market success from formulation. It’s intended to be sort of a food scientist’s best friend in terms of new product ideation."
NECTAR’s Mission: Democratizing Sensory Data
At its heart, NECTAR is committed to compiling the most comprehensive public dataset detailing how omnivores experience the taste of sustainable protein products. This mission is not merely academic; it’s driven by a pressing need to accelerate the transition away from conventional meat production, a significant contributor to environmental degradation. By making this data and the resulting AI tool openly accessible, NECTAR aims to democratize the innovation process within the food industry.
"Basically, Nectar is trying to amass the largest public data set on how sustainable protein products taste to omnivores. That’s what we have set out to do," Cotto stated. "We’re building that, and we are working heavily with academics to operationalize that data."
The partnership with Stanford’s computer science department has been instrumental in translating NECTAR’s extensive sensory data into actionable insights. The collaboration has focused on developing a sophisticated AI model that can not only predict sensory outcomes but also correlate them with market success.
The Technical Blueprint: A Multi-Modal AI Approach
From a technical standpoint, the AI model NECTAR is developing is built upon the foundation of Large Language Models (LLMs). The team is fine-tuning these models to excel at sensory prediction. This is not a singular approach but rather a multi-modal strategy, incorporating various data streams to achieve a holistic understanding of food perception.
"I am not an AI scientist myself here, so we are heavily partnered with Stanford and their computer science team, but it is an LLM base," Cotto elaborated. "We’re basically fine-tuning an LLM to be able to do this sensory prediction work, and it’s a multi-modal approach. There’s a similar project that’s been done out of Google DeepMind called Osmo for smell and olfactory, and we’re working with some of the folks that worked on that in order to model taste and sensory more broadly, and then connect that to sales outcomes."
This innovative approach draws parallels to Google DeepMind’s "Osmo" project, which focused on modeling smell and olfactory senses. By adapting and expanding upon such methodologies, NECTAR aims to create a robust system capable of predicting taste and overall sensory experience, and crucially, linking these attributes to commercial viability.
The Bezos Earth Fund AI Grand Challenge: A Phased Approach to Innovation
The Bezos Earth Fund AI Grand Challenge for Climate and Nature is a significant initiative, allocating a total of $30 million to a select group of projects. The challenge operates in distinct phases, designed to foster the development and refinement of groundbreaking AI solutions.
"It’s the Bezos Earth Fund AI Grand Challenge for Climate and Nature. It’s $30 million going to these projects. There were 15 phase two winners that each received $2 million and have to deliver over two years," Cotto explained.
The initial Phase One grant, a more modest $50,000, served as a crucial incubator for ideas. NECTAR utilized this period to meticulously prepare its submission for Phase Two. Over six months, the team focused on integrating NECTAR’s proprietary sensory dataset with sales data to identify which sensory attributes most strongly predict sales success. Concurrently, they worked on connecting the sensory data to molecular-level ingredient datasets. The ultimate vision is a predictive chain that can forecast sensory outcomes from ingredient lists alone, and subsequently, link specific sensory characteristics to market performance.
Rigorous Sensory Testing: The Bedrock of NECTAR’s Data
The foundation of NECTAR’s AI model is its extensive and rigorously collected sensory data. This data is gathered through in-person, blind taste-testing sessions designed to mimic authentic dining experiences while eliminating bias.
"It’s all in-person blind taste testing," Cotto detailed. "In our most recent study, we tested 122 plant-based meat alternatives across 14 categories. Each product was tried by a minimum of 100 consumers. They come to a restaurant where we’ve closed down the restaurant for the day, but we want to give them that more authentic experience. They try probably six products in a sitting, one at a time, and everything is blind, so they don’t know if they’re eating a plant-based product or an animal-based product and then they fill out a survey as they’re trying the product."
This meticulous methodology ensures that consumer feedback is based purely on the sensory attributes of the product, free from preconceived notions about its origin or category.
Expanding Data Horizons: From Meat Alternatives to Dairy
NECTAR’s data collection efforts are ongoing and expanding across various alternative protein categories. The annual "Taste of the Industry" survey provides a snapshot of the market, with testing volumes increasing year over year. In 2024, approximately 45 plant-based meat products were evaluated, a number that surged to 122 in 2025.
Beyond meat alternatives, NECTAR is actively exploring emerging sectors. Their research has delved into "balanced protein" or "hybrid products" that blend animal meat with plant-based ingredients, microproteins, or savory vegetables. Nearly 50 products in this category have been tested.
Looking ahead to 2026, NECTAR is set to shift its focus to dairy alternatives. The upcoming "Taste of the Industry" report will feature an evaluation of 100 dairy alternatives across 10 distinct categories, with findings slated for release in March. This broad scope underscores NECTAR’s commitment to providing a comprehensive understanding of the alternative protein landscape.
The Taste-Sales Nexus: Unveiling Market Drivers
A pivotal aspect of NECTAR’s research is the correlation between sensory performance and sales data. While NECTAR’s primary focus remains on sensory evaluation, the organization is keen to answer the fundamental question: do better-tasting products translate into higher sales?
"The Nectar data set is mostly just focused on sensory. That’s the core of what we do," Cotto stated. "We are also interested in answering the question ‘do better-tasting products sell more?’ In our last report, we conducted an initial analysis of overlapping sensory data with sales data, finding that better-tasting categories capture a greater market share than worse-tasting categories. Better-tasting products are capturing greater market share than worse-tasting products. In certain categories, that seems to be agnostic of price. Even though the product might be more expensive, if it tastes better, it is capturing a greater market share."
This preliminary analysis suggests a strong positive correlation between superior taste and market success, even in instances where price points are higher. However, the team acknowledges the complexity of this relationship, recognizing the influence of numerous confounding variables such as marketing spend, store placement, and shelf positioning.
To address this, NECTAR is actively collaborating with data providers to obtain more granular sales data. They are leveraging the Nielsen consumer panel, a vast repository of grocery store transactions that tracks household purchasing behavior over extended periods. This allows NECTAR to connect detailed sensory profiles with actual consumer purchasing decisions, aiming to identify the specific sensory elements that drive sales.
Deconstructing Flavor: The Role of Molecular Data
The integration of ingredient lists and molecular data into the AI model is another critical component of NECTAR’s strategy. The food product development landscape often involves "black boxes," particularly concerning flavor formulations. Companies typically keep their precise formulations confidential, making it difficult for external researchers to gain visibility.
"We’re trying to say, there are a lot of black boxes in food product development because flavors are a black box," Cotto explained. "We don’t have a lot of visibility into companies’ actual formulations. We’re trying to determine if we can extract publicly available information from the ingredient list and identify the molecular-level components of those ingredients, and then determine if any correlations can be drawn between them."
By analyzing publicly available ingredient information and identifying the molecular components of those ingredients, NECTAR aims to uncover correlations that can predict sensory outcomes. This, combined with image data of products, forms a comprehensive input for the AI model, enhancing its predictive capabilities.
The Ultimate Deliverable: An Open-Source Toolkit for Innovation
The overarching goal of the two-year grant is to deliver a practical, open-source tool that the entire food industry can utilize. This tool is envisioned as a powerful aid for food scientists and product developers, enabling them to navigate the complex landscape of product creation more efficiently.
"The idea is to deliver an open source tool for the industry to use," Cotto stated. "The goal would be that you can put in all the constraints you have for sustainability, cost, nutrition, and demographic need, and that it would help you get to an endpoint where you don’t have to do a bunch of bench-top trials and then expensive sensory."
This tool promises to significantly reduce the time and cost associated with bringing new alternative protein products to market, thereby accelerating the adoption of more sustainable food choices.
Navigating Data Privacy and Industry Collaboration
The commitment to open-source development raises important considerations regarding data privacy and how companies will engage with such a tool. NECTAR emphasizes that its intention is not to acquire proprietary formulations but rather to provide a platform for companies to enhance their own product development processes.
"Data privacy is a big thing in this space. We don’t have any interest in companies sharing their proprietary formulations with us," Cotto clarified. "The goal is that they would be able to utilize this tool, download it to their own servers, and put in their private information and use it to make better products. If we’re rapidly increasing the speed at which these products come to market and they are actually successful, that would be a success for us."
NECTAR also acknowledges the existence of other initiatives in the alternative protein space, such as those by NotCo and the Institute of Food Technologists (IFT). While recognizing the similarities in objectives, NECTAR distinguishes its approach by its proprietary sensory dataset and a clear mission-driven focus.
"I think everybody is trying to do similar things, but with slightly different inputs and different approaches," Cotto noted. "We are open to collaborating and learning from people. Our end goal is a mission-driven approach here, not to make a ton of money, so it depends on whether or not those partners are aligned with that goal."
NECTAR’s methodology, centered on its extensive sensory data, offers a unique perspective compared to models trained on academic literature. While the IFT model, for instance, draws from published papers, NECTAR’s AI is specifically tailored to understand the nuances of consumer perception.
The Larger Climate Imperative
Ultimately, NECTAR’s work is deeply intertwined with a broader climate mission. The organization recognizes the urgent need to reduce global meat consumption to mitigate the impacts of climate change. However, they believe this transition cannot be solely achieved through appeals to consciousness; it requires making sustainable protein options genuinely appealing.
"Nectar’s specific directive is, how do we make these products favorable and delicious?" Cotto concluded. "We know that we need to reduce meat consumption in order to stay within the two degrees of climate warming, and we’re not going to get there by just telling people, ‘eat less steak.’ We have to use that whole lever and make the products really delicious so that people will be incentivized to buy them more and reduce consumption of factory-farmed meat."
By developing AI-powered tools that facilitate the creation of superior-tasting alternative proteins, NECTAR is not just advancing food science; it is actively contributing to a more sustainable and environmentally conscious future for food production. The $2 million grant from the Bezos Earth Fund marks a pivotal moment, empowering NECTAR to accelerate this critical transformation.
