ProfilePrint is an AI-driven predictive and prescriptive food profiler. Its patented food fingerprinting technology provides users with accessible, affordable and portable solutions to predict outcomes at the point of use, within seconds. ProfilePrint’s solution allows sellers and buyers to ascertain the agreed quality of food ingredients digitally, expediting existing assessment processes and facilitating trade – from the farm all the way to end-buyers.

ProfilePrint is a Singapore-based company with a global footprint. For more information, please visit https://profileprint.ai/.

ProfilePrint is an AI-driven food profiler, with a patented food fingerprinting technology that allows users to predict outcomes within seconds.

Additional angles/information:

• ProfilePrint is a rapid food ingredient analysis solution that leverages on AI to provide an objective snapshot of the quality of a food ingredient, within seconds
• ProfilePrint’s patented AI food fingerprinting technology empowers any untrained user to ascertain the quality of a food ingredient like an expert

We are an agri-food tech company based in Singapore, with sales offices in Belgium, Japan, and soon to open in China. Founded in 2017, we are funded by the world’s largest food ingredient conglomerates (such as Cargill, Olam, LDC, Sucafina and Sinar Mas), international venture capital funds and the Singapore government.

Natural ingredients have unique molecular signatures – even between origin, condition, climate etc. We capture this by exciting the molecules with short wavelength electromagnetic spectrum, and digitise this information in the form of a “digital food fingerprint”.

ProfilePrint’s patented AI food fingerprinting technology is able to synthesise complex interrelated parameters and sensory data into a single digital fingerprint, allowing users to rapidly predict quality and profile of a food sample within seconds, after they have built their AI models.

ProfilePrint expedites existing quality assessment processes and offers sellers and buyers a common digital reference point, expediting existing quality assessment processes and facilitating trade. It is:

• Affordable (compared to laboratory testing and human panels)
• Rapid
• Flexible (different models can be built for multiple verticals)
• Non-destructive
• Portable and accessible

• ProfilePrint combines both sensor technology and a SaaS platform to assess a food ingredient at any point in the supply chain
• The portable analyser captures molecular signatures of food ingredients, and this data is processed using our proprietary AI technology in the cloud to generate predictions

It is a 3-step process:

1. First, our analyser captures the molecular signatures of the food ingredient;
2. Second, we process this data using our proprietary algorithm stored in the cloud to build the best model that can generate the most accurate predictions for your specific use case;
3. Third, this prediction is generated as a digital report for clients

For example, if you are building a sensory model for coffee beans, then we train this model with coffee samples and data from your existing assessment methods, like expert Q-graders. The more data we have to train the model, the more robust it will be. When we apply this model with an unknown coffee sample scanned by our analyser, it is able to predict the sensory profile of this coffee without human tasting.

ProfilePrint leverages on AI to process both small data of the product such as its molecular signature, and big data such as origin; weather; altitude etc. to generate insights and predictions.

When it comes to sensory predictions, words like “bitterness”, “astringency”, “umami” are descriptive words that may not mean the same to different tasters  We deploy AI to recognise patterns based on data that have been provided by the client (for example, from their master tasters). With sufficient training data, the AI can recognise patterns to help generate a prediction based on how they correlate to the actual raw ingredient’s molecular signature. 

ProfilePrint can incorporate critical control points into the monitoring process so clients can apply corrective action if problems arise at any stage of the supply chain from sourcing, manufacturing to fulfilment.

ProfilePrint is currently focused on coffee, cocoa, tea, grains and oilseeds. ProfilePrint can technically be used for all ingredients as long as the samples have a molecular signature difference.

ProfilePrint currently has four main solutions/modules:

1. ProfilePrint Hub (specific to coffee industry, no need to mention if not relevant to the client): AI-powered cupping reports for green coffee beans, according to SCA’s grading protocols.
2. Classify: Rapidly classify food ingredients based on its signature profile. E.g. pass/fail, grades, country of origin etc.
3. Profile: Ascertain multiple complex and interrelated parameters, including sensorial parameters, chemical or physical attributes such as moisture, density, fat etc.
4. Blend: Suggest match or a blend of ingredients to achieve the required profile

ProfilePrint Hub allows anyone to be able to grade green coffee beans using ProfilePrint.

ProfilePrint generates a digital report of green arabica coffee beans based on a model trained by expert Q-graders, and according to SCA’s green coffee grading protocols. This includes cupping scores and other sensory profiles, all without the need to roast, grind or cup.

With just 50g of samples, sellers can receive their ProfilePrint cupping reports that can be shared digitally with their buyers globally. Buyers are also able to more confidently purchase specialty arabica beans scored by ProfilePrint’s AI Q-grader model.

Currently, we have launched a model for green arabica coffee beans, according to global standards set by SCA, and trained by expert Q-graders. We are working on models for other verticals, such as cocoa and rice.

ProfilePrint’s HQ is in Singapore, and we have sales offices in Belgium, Japan and China. Our systems are currently deployed in 26 cities across 5 continents (North America, Latin America, Africa, Europe, and Asia).

Note: These problems may not be the same for every one, but in general:

Expensive and onerous process

• Currently: Current methods of quality assessment vary from laboratory testing which is expensive and time consuming, to human tasters, which is onerous and subjective
• With ProfilePrint: ProfilePrint’s technology is affordable, rapid, and accessible

Unfair pricing

• Currently: At the producer (or farmer) level, crops are visually graded, and this subjectivity can result in unjustified price point for farmers. In addition, some farmers are also paid based on volume regardless of crop quality, which does not incentivise them to improve farming methods. Farmer poverty continues to worsen despite increasing demand for premium quality and continuous improvement of farming methods
• With ProfilePrint: ProfilePrint allows producers greater insight and clarity into the quality of their beans, allowing them to set fairer prices for their crop. Similarly, buyers also have a better understanding of the product prior to purchase, and can be more assured of its quality and consistency, allowing them to more confidently make their purchase

No common standard

• Currently: Often, multiple rounds of sampling is required before a decision is made. Even after a purchase, buyer and seller may still disagree that it was the quality that had been agreed upon earlier, leading to potential wastage
• With ProfilePrint: ProfilePrint’s rapid food ingredient analysis solution provides an objective snapshot of the quality of a food ingredient. ProfilePrint is transforming the food ingredient supply chain by digitalising the entire food sampling process, allowing users to search for, or grade, food based on its molecular signature without physical samples.

• Food verticals that have high velocity and highly differentiated products that require rapid quality assessment
• Key stakeholders along the entire supply chain (i.e. manufacturers, traders, exporters, cooperatives, farmers/producers)
• All sizes of companies – from food conglomerates to MNCs, SMEs and start-ups

As of 2022, we serve some of the world’s largest food conglomerates, including Cargill, Louis Dreyfus Company (LDC), Olam, Sinar Mas and Sucafina, who are also our strategic investors since our last funding round in August 2022.

We also serve global food manufacturers as well as mid-sized traders, cooperatives, and farmers.

We have various solutions to cater to each client’s needs:

1. For food companies, manufacturers etc.: ProfilePrint offers a subscription-based model

• Standard rate is US$15,000 per year on a subscription basis, which will include rental of the ProfilePrint analyser and full access to the modules available on the SaaS platform (classify, predict, blend)
• Model building and customisation costs an additional US$5,000 to US$10,000 per model
  • In general, it can cost between US$15,000 to US$30,000 depending on how much customisation and support is needed
  • Why do we run on a subscription-model? This allows us to offer clients the most updated versions of our solution (at a small fee, if at all)

2. For green coffee bean farmers and producers:
   • Pay-per-use at US$10 per report
   • Partners (cooperatives etc.) at US$5 per report

Traditionally, to sell internationally, producers will need to send the cupping score of their beans to their buyer.

Producers will need to either cup their beans personally, which to buyers, may not be reliable in terms of cupping score, or to send it for lab testing, which is expensive at about US$50 per report, which will take 2 to 3 weeks.

ProfilePrint has a global SCA model trained by expert Q-graders, which includes cupping scores and other sensory profiles, all without the need to roast, grind or cup. With just 50g of samples, sellers can receive their ProfilePrint cupping reports that can be shared digitally with their buyers globally. This is affordable at US$10 per report, and takes just 5 working days upon receipt of beans in Singapore. Buyers are also able to more confidently purchase specialty arabica beans scored by ProfilePrint’s AI Q-grader model, especially as we act like a neutral party.

Beyond the immediate operating efficiencies and cost-savings we bring to the supply chain, ProfilePrint drives sustainability in a big way. Our technology provides traceability features allowing end buyer to access the digital signatures as well as geographical and harvest records of the ingredient they consume. It alleviates farmer poverty by ensuring that they are paid more fairly based on quality and ultimately providing more d objectivity in the industry.

Also, carbon footprint is significantly reduced now that we require a smaller amount of ingredients for assessment (e.g. 50g instead of 500g for green coffee beans) and with the reduced need for shipping samples, roasting and human tasting, reducing carbon footprint.

• We are currently applying for ISO27001 to enhance our cyber security posture
• We do not share client-data. Each clients have their own dedicated and exclusive access to their own data