Science + Technology

We have spent over 8 years and 250,000 human hours of effort in creating our platforms and program. Our mission: Is to democratize more than 25+ years of scientific research in micronutrition, genetics, exercise, sleep, fasting and other lifestyle factors to create an evidence informed software platform for preventive health. Use cutting edge technology: hardware, software, mobile app to deliver precision programs with the ability to track, measure and improve health outcomes on a continuous basis. Our gen 7 platform & program delivers the best-in-industry outcomes to optimize your key nutritional status and disease risk markers, that are known to impact your health, longevity and healthspan.

Clinical Evidence Stratification System

Our team of scientists(expertise in Developmental Biology, Medical Biochemistry, Biotechnology, Genetics, Molecular Biology, Neuroscience and Cell Biology) have spent over five years developing a de-novo science system, by meticulously re-evaluating and stratifying all human clinical trials(worldwide sources: PubMed, Cochrane), conducted on micronutrients and botanical interventions done in the past 25 years.

This proprietary Science system categorized the clinical evidence, i.e 40k+ studies conducted across 1000+ active ingredients, 4000+ health conditions and 60+ key health benefits, thereby quantifying the depth of research done in this space over time, and establishing a hierarchy of effectiveness for specific health interventions. This includes weighting evidence from high quality studies(RCTs) to determine the most efficacious formulations and dosages for specific health interventions, that then forms the starting basis of evidence-based nutrient recommendations we provide our users. Which is then further personalized based on the user's age, gender, BMI, health history and genetic data.

Our goal with this system is to demystify and democratize complex science around health and nutrition. And with this transparent and complete access to our well qualified science system, empower our users with the scientific knowledge on any/all nutrient interventions/supplements that are part of their Project Serotonin program recommendations, or for what they may be taking, or actively considering for their unique needs.

Studies Examined
Ingredients Examined
Health Benefits Examined
Health Conditions Examined

Continuous measurements and optimization is central to our Project Serotonin program. The science system also supplies crucial data for the ongoing qualitative and quantitative health impact/outcomes measurements (HIMs i.e. Blood biomarkers) - as employed in the clinical trials, to evaluate the real-world effectiveness of ingredients in an ongoing basis.

Nutrigenomics System

Supplements offer a boost to our well-being, but their effectiveness is heavily shaped by our genetic composition. The human genome harbors genetic variances that exert a profound influence on how we metabolize nutrients. For instance, key genes such as MTHFR, VDR, COMT, GSTM1, and GSTT1 play important roles in metabolizing nutrients/neurotransmitters, and detoxification. Variants in these genes affect how our bodies process vitamins, minerals, botanicals, and other supplements. Thereby directly impacting the quantum of benefits or lack thereof that one might observe/experience with these supplements.

The burgeoning discipline of nutrigenomics allows us to tailor supplement programs based on the user's unique genetic blueprint. This entails genetic testing to unveil specific single nucleotide polymorphisms (SNPs) associated with nutrient metabolism. Subsequently, tailored supplement recommendations are designed, accounting for an individual's genetic proclivities, dietary habits, and health objectives. Regular monitoring and adjustments ensure that the regimen evolves in harmony with changing needs.

To attain such a high degree of personalization, we utilize genetic variants sourced from PubMed and the GWAS catalog. We've established a meticulously curated database, overseen by experts, housing the most influential SNPs supported by substantial clinical study evidence. This system is organized according to clinical conditions and genotypes proven to influence specific traits. This approach holds great promise and stands poised to elevate individuals' well-being by enabling a laser-focused approach to targeted supplementation protocols. By identifying the precise genetic variants linked to specific health conditions or traits, we can suggest supplements that are more likely to be advantageous for that particular individual. This contrasts starkly with the conventional, one-size-fits-all method of supplementation.

Genomics-based personalization, through the use of genetic variants from our curated system, holds the potential to revolutionize our understanding of the effectiveness of supplements and promise of markedly improving individuals' health and vitality.

Genes Evaluated
Genes Evaluated
SNMS Evaluated
Traits Evaluated

Product Qualification System

"Cultivating Confidence in Dietary Supplements: Project Serotonin's Commitment to Quality and Personalized Wellness"

Popular dietary supplements can pose potential risks, including subpar manufacturing, allergic reactions, and a failure to meet dietary requirements. These risks further lead to inadequate dosages, substandard formulations, and inactive ingredients that offer no clinical benefits.

Reputable brands adhere to stringent labeling practices, good manufacturing standards, and quality regulations. However, customers may overlook or underestimate the significance of that quality information. Project Serotonin employs a proprietary supplement knowledge system to identify products from esteemed brands that meet quality standards, such as Good Manufacturing Practices (GMP), USDA Organic, Kosher, USP, and UL testing, all subject to scrutiny by regulatory authorities. The recognized brands provide comprehensive product labeling, disclosing details about fillers, colors, sweeteners, and more, promoting transparency and building trust with consumers.

Products Qualified
User Scanned

Allergens can lurk in various supplement products, and most users are unaware of the potential allergic reactions they may trigger. Even a small amount of allergen can lead to hypersensitivity. Project Serotonin aims to educate users about the nine major allergens—sesame, milk, eggs, fish, crustacean shellfish, tree nuts, peanuts, wheat, and soybeans.

In recent years, there has been a concerted effort to improve allergen labeling in dietary supplements, making it easier for individuals with allergies to avoid products that may contain allergens. Project Serotonin takes user allergen inputs seriously and carefully assesses it against the nutrient support stack using the proprietary supplement knowledge system.

Project Serotonin's proprietary supplement knowledge system facilitates the curation of intelligent nutrient stacks that exclude unnecessary sugars and talc, while accommodating users' dietary preferences, whether they follow a vegan, vegetarian, or mixed diet.

This system ensures the optimal dosage, formulation, and active ingredients in selected dietary supplements, delivering personalized clinical benefits tailored to an individual's genetics, comorbidities, age, gender, body weight, lifestyle, diet choices, and any other restrictions.

The Algorithm

Our algorithm includes 600 parameters per user profile, periodic blood test based assessments, and our knowledge system-derived inputs to build personalized supplement stacks to address many health objectives. In addition to nutrient optimization, we also address health parameters like HbA1c, homocysteine, hs-CRP and ApoB.  Project Serotonin also has algorithms for specific health goals like lean muscle mass maintenance, VO2max, stress relief, women’s health, men’s health, and sleep support. 

Cohesive Evaluation of Diverse Health Metrics: Optimization of serum nutrient levels and levels of specific key health markers is the primary goal of the Serotonin algorithm. Anthropometric parameters like weight, BMI and body composition metrics are considered to understand nutritional needs of an individual. Blood biomarker readouts are then factored in to customize the nutritional inputs to the present, measured health, and nutritional status of an individual. 

Additionally, genetic factors that determine the predisposition of an individual towards nutrient sufficiency or health parameter dysfunction are then factored into the algorithm for every profile. Weightage of evidence from published clinical trials is used to fine-tune choices of dietary supplement ingredients, formulations and dosages that are required for every profile. Inputs from personal health history and family health history are also considered in the choice of formulation and dosages of nutrients that are chosen for an individual. Healthy lifestyle practices such as exercise performance patterns, and sleep patterns are also factored in with appropriate weightage for the choice of nutrients and herbals for specific health aspirations and nutrient sufficiency status. We also account for nutrient depletion resulting from leisure lifestyle habits such as smoking, alcohol consumption, and occasional recreational substance usages.

Most importantly, we evaluate the negative interactions of supplements with prescription medications and eliminate all choices that can lead  to even the mildest of adverse events. This ensures high safety of the program stack.

Personal choices like diet, food allergies and other consumption preferences are considered to finally translate the ingredient, formulation and dosage choices into the most accurate supplement product fit from the ones that are available in the market. Evidence from scientific studies is also used to create a precise regimen for synergistic consumption of dietary supplements, in order to maximize the beneficial effects on human health. 

Our algorithm is brand agnostic, and therefore we create heterogenous stacks using products from many high quality brands that suit the health needs of every individual. 

Periodic assessments of human health & continuous longitudinal monitoring are therefore at the heart of the Project Serotonin program and algorithm. Our algorithm is a novel, progressive, iteratively learning system that accommodates the changing volume of clinical evidence, anthropometric measurements, user experience, clinical readouts and genetics to create new personalized nutrient inputs for every user at every stage of the programThis ‘Adaptive Integration’ mode of the algorithm makes it possible to include more knowledge inputs such as AI-enabled analytics, metabolomics  and other knowledge systems in the future.

The algorithm is version controlled and updated every three months to include learnings from data analysis and therefore delivers highly precise outputs for each user in every phase of the program. The algorithm is able to simultaneously analyze weighted inputs from multiple user parameters, fitness bands, product market as well as updated knowledge from genomics and clinical trials. The data hierarchy used in the Serotonin algorithm has been independently validated by data models suggested by an AI/ML system as well.

ApoB: A Critical Marker of Cardiovascular Health

Apolipoprotein B (ApoB) is a protein that is found on the surface of lipoproteins, which are particles that transport cholesterol and other fats in the blood. ApoB is the main structural component of atherogenic lipoproteins, which are lipoproteins that can contribute to the development of atherosclerosis, a disease characterized by the buildup of plaque in the arteries.

hsCRP: A Critical Marker of Inflammation and Cardiovascular Risk

High-sensitivity C-reactive protein (hsCRP) is a protein produced by the liver in response to inflammation. Inflammation is a normal bodily response to injury or infection, but chronic inflammation can lead to a number of health problems, including cardiovascular disease.

Understanding Your Iron Status

Iron is an essential mineral that must be consumed in our diet to ensure optimal functioning of our bodies. It serves as a fundamental component of both hemoglobin and myoglobin and plays a pivotal role in a diverse array of vital metabolic processes.


Methylation is a chemical process that adds a methyl group (CH3) to biological molecules. It is a dynamic and critical cellular process that plays a role in a variety of functions, including gene expression, protein function, and DNA repair.  

Personalizing Wellness: Harnessing AI and ML for Optimal Health

The human body is the result of code that is encoded in the DNA. There are four variables (AGTC) that are strung in succession, three billion times. 99.9% of this code is identical amongst humans with variation in just about 0.1% of the code. It is this variation that makes each of us unique, not just in our appearance but also in the way we respond to food, medication, stress, our environment and thereby how they impact our health status. While the influence of some of these factors are known, conventional science has a long way to go to develop a better, more deeper and comprehensive understanding.

Additionally, we also differ from one another on our physiological and behavioral profile(parameters like age, diet preferences, laboratory results, lifestyle, alcohol consumption, smoking, exercise, and sleep etc). This increases an individual’s uniqueness or the variation in a population. The total number of combinations that are possible (Genetic + Profile + Behavioral) cannot be factored in a manual assessment, even when conducted by highly trained healthcare specialists.

Thus, most of the current supplementation/lifestyle recommendations use a few aspects at the profile level (age, gender, weight, lab results) to devise protocols. Given how broad brushed the current systems are, only a very few respond to the prescribed protocols in the desired manner. While widely accepted and practiced, the level of personalization in this method is almost non-existent.

At Serotonin, we use real world data from our users to build models using Machine Learning and AI technology to recommend hyper-personalized programs and regimens unique to the user. These models factor in a user’s genetics, profile and behavioral parameters, using 800,000+ data points of each individual user in order to ascertain what is required to improve and maintain their optimal health status, and mitigate disease risk.

From our analysis we know that this combination of data points can significantly increase the accuracy of the models. An example of this is shown below.

These three figures show the best models created for Serum Vitamin D improvement prediction with supplementation. These models help personalize the effective dosage, formulation and specific product(brand) of the recommended supplement. Fig 1 shows the best model for predicting Serum Vitamin D using profile data only, Fig 2 uses genetic data only, and Fig 3 uses both genetic and profile data. All three models used the same training data set and test data for comparability of the models. The accuracy of the models is based on the Weighted Absolute Percentage Error (WAPE), that measures the overall deviation of forecasted values from the observed values. The most accurate model is highlighted by the lowest WAPE score (0.16) - which uses both profile and genetic data.

ML Model using profile + genetic data

There are many use cases for similar models with respect to health and wellness, including supplement - dosage, formulation and regimen, exercise - type, intensity, and duration, sleep - time and duration, fasting - time and duration and so on. Our models have the ability to give the user precise recommendations and on a real time basis. Above all, the models are based on closed loop learning systems and this constantly improves the accuracy of the models as more data flows in from the user.

ML Model Using Only Profile Data

There are many use cases for similar models with respect to health and wellness, including supplement - dosage, formulation and regimen, exercise - type, intensity, and duration, sleep - time and duration, fasting - time and duration, and so on. Our models have the ability to give the user precise recommendations and on a real-time basis. Above all, the models are based on closed loop learning systems and this constantly improves the accuracy of the models as more data flows in from the user.

ML Model using only genetic data

There are many use cases for similar models with respect to health and wellness, including supplement - dosage, formulation and regimen, exercise - type, intensity, and duration, sleep - time and duration, fasting - time and duration and so on. Our models have the ability to give the user precise recommendations and on a real time basis. Above all, the models are based on closed loop learning systems and this constantly improves the accuracy of the models as more data flows in from the user.

Smart Tray Hardware

The Serotonin Hub hardware is designed to be the centerpoint of your daily supplementation experience. Tucked seamlessly into your supplement box, the Hub is the point at which you tell us what you are taking, and where we provide instantaneous, actionable feedback about your daily supplements.

Interacting with the Hub is simple - just bring the tag on your supplements or other products to the center of the scanning pad and lower them to the surface. The front of the surface will light up in green to confirm that the scanner is active - followed by a double beep and green flash to confirm when supplement consumption has been logged.

The position of the hub is adjustable - with multiple depth settings available to support easy scanning of the products featured in your unique stack.

The onboard battery in the Hub will last for months, and when it is time to charge, you’ll receive a notification on your phone via the Serotonin app. You don’t have to worry about keeping track of a special cable to charge the Hub, any USB-C charger will do the trick.

1. High Speed NFC Front End

Lightning fast supplement scans are enabled by our deeply integrated NFC front end and speed optimized firmware.

2. USB-C Recharging

Any USB-C charger will work to refuel the 2000 MaH battery embedded in the Smart Tray.

3. Proximity Sensor Based Wakeup

The proximity sensor at the center of the Pad quickly wakes up the NFC scanner when something comes close, regardless of light conditions.

4. Nordic NRF52832

The Nordic NRF52832 System-On-Module at the heart of the Smart Tray ensures quick interactions and seamless Bluetooth interactions with your phone, including over the air firmware updates.

5. Dynamic UI

A ring of bi-color LEDs and a tiny speaker confirm your interactions with the Smart Tray.

6. Custom BLE protocol

We developed a custom Bluetooth low energy protocol to maximize the flexibility of the Smart Tray. During pairing or firmware updates, the Smart Tray and phone operate in a symmetrical, paired connection. During normal daily use, the device switches to a Bluetooth advertising mode, chirping out codes representing whatever supplement consumption events are queued, until your phone confirms that it is synchronized. This helps maximize the battery life of the Smart Tray, while ensuring that your supplementation history is faithfully shared with the app.

7. CNC Machined Aluminum Enclosure

The ultra-slim enclosure of the Smart Tray tracker consists of two CNC machined and anodized 6061 Aluminum components, and three glass filled resin 3d prints, manufactured in over 1000 tiny layers.

Gen 1 Tracker

We settled on an accelerometry based model for tracking - using a recurrent neural network written in low level code to make probability judgements on the likelihood that any series of movements represented a use event. Our Gen 1 Tracker was the first generation of hardware to enable this approach, with a back that slid off for a user replaceable battery, and replaceable adhesive mounts to carry the tracker forward as bottles were emptied.

Gen 2 Tracker

Learning from Gen 1 of the Tracker, we created Gen 2 - with an easy to replace battery module, and refined injection molded enclosure, the Gen 2 Tracker was the first to be deployed to all Serotonin users.

Gen 3 Tracker

To reduce the waste associated with bottle changeovers, and to create a unified experience between hardware and app, we transitioned to an NFC based tracking model. Generation 3 of our tracker was designed to support this experience, as standalone hardware with a dedicated area for the phone to be positioned during use.

Gen 4 Tracker

To more deeply integrate the scanning and consumption experiences, we made a new version of the NFC scanner that is unified with the tray in which users receive and store their supplement stack, transforming the tray into a Smart Tray.

By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.