Embracing AI To
Enhance Human Connection

At Cloverleaf, our mission is to unleash people to do their best work. We do this by using automated coaching to empower better connections across teams and organizations, providing you with the tools to communicate, motivate, and understand each other in actionable ways. We’re embracing AI to ensure your coaching tips are accurate, practical, and research-backed. No fluff or filler, just 100% coaching tailored for you.

Certain features of Cloverleaf, like Insight Search and the Reflection prompts on your coaching tips, leverage AI-powered building blocks. These building blocks enable Cloverleaf to connect your searches to the most relevant tips to improve the quality of your Coaching Experience. Every tip, prompt, nudge, and assessment result is scrutinized, vetted, and audited by our team of certified coaching consultants and researchers. You can rest assured that it’s still Cloverleaf under the hood.

We have extremely high standards for data privacy and fully anonymize inputs into these AI models by default and explicitly seek your permission when this is not possible. The only person who can ever access and read this data is you.

To learn more about Cloverleaf’s data privacy, trust, and ethics policy, click here.

The world is your oyster with AI. Currently, we only leverage AI to increase the speed and quality of tip delivery. As we develop new features, we’ll keep you in the loop and will always clearly outline how AI is used in our product. Your feedback is always welcome, and we’d like to thank you for going on this journey with us.

If you have questions about how Cloverleaf uses AI in our product, or in general, read our Frequently Asked Questions below.

In short, the answer is yes. While “AI” is a broad term encompassing various technologies, it’s crucial to clarify what we at Cloverleaf mean when we say “AI.” We define AI as the use of software algorithms that function and evolve based on training data, which updates the weights of a model. This differs from non-AI software that operates on explicit rules written in code. Like many tech companies, Cloverleaf integrates both rule-based algorithms and weight-based models to enhance our product. As new technology and techniques emerge, we rigorously study their potential and constraints, and strategically incorporate these advancements to further our ultimate mission: “Unleash people to do their best work.”

We’re always working to improve our product which means adding AI features when and where it makes sense. Here’s a list of the current Cloverleaf features that leverage AI. We’ll keep updating this list when we add new things. Keep an eye out for The Cloverleaf Compass and our Product Release notes to stay updated.

Insight Search: Insight search is powered in part by a text embedding model. Text embedding is the process of transforming words or phrases into numerical vectors that represent their semantic meaning. It allows us to connect your free text searches with the most relevant Cloverleaf coaching tips. We make sure your searches are private by removing any personal information, like names. We review these anonymized searches to see where we can make improvements, but we don’t use them for AI training.

Tip Generation: The tips you see on Cloverleaf are rooted in research and today are written by real people who work for Cloverleaf. As AI technology improves, we’re using large language models (LLMs) to help with things like checking spelling and grammar. We’re also using it to improve our tip quality and tone at scale. We’ve made an intentional decision in our approach to ensure that 1) generative AI does not use any customer data and 2) any content modified by AI tools is truthful and rooted in research.

We use generative AI tools like ChatGPT as a writing tool, not a writing replacement. We employ a team of coaches, assessment consultants, and researchers whose combined expertise spans both the theoretical and applied psychometric evaluation space. We’ve spent years refining our writing process to be able to produce coaching tips that feel personalized and “spot on”, and are now able to further scale our robust, research-backed process with emerging generative AI tools.

There are two types of mistakes that can emerge from AI systems, those that result from accidental errors or “bugs” in the software and those that are a result of the model’s inaccuracy. For mistakes emerging from bugs, we treat this similarly to any other defect in our product and work to mitigate the impact quickly, recover data loss if applicable, and communicate any impact to exposed customers. The second class of mistakes that are a result of model accuracy requires an entirely different mindset. Each integration of AI-powered software into our product requires a cohesive design that captures all possible potential failure scenarios. We build these guardrails into the product prior to public release and internally stress test them to minimize the negative impact of model inaccuracy.

  1. Insight Search: Our search product retrieves content from our library of tips and leverages AI to match a user’s query to the best tip. AI does not generate our responses, but we have built-in guardrails to block offensive and harmful queries and handle biases with care.
  2. Tip Generation: As mentioned above, we leverage generative AI tools to assist with scaling and we do not use generative AI technology as the ideation engine. This means that our content is all rooted in research that has been sourced by human experts in the field.

This intentional approach allows us to mitigate mistakes typical of AI systems while delivering the most value to our customers.

The only customer data that we feed into our AI systems is anonymized query data from insight searches. Each query is stripped of all identifiers that tie it back to the authoring customer. We also strip out all names recognized by our system before feeding them into our AI system. This removes all identifiable information that would link a query back to a customer. We store these queries in an encrypted way that only allows the application to operate on them, which allows customers to retrieve their search history. Our AI system will also use these queries as a signal to deliver more relevant tips as part of your daily coaching. Customers have the ability to disable all query history, which will disable all of these features. Cloverleaf will never share customer search queries with their organization.

In the future, we hope to leverage AI and Machine Learning to make the Cloverleaf increasingly personalized, and therefore increasingly valuable, for each customer’s experience. This may involve leveraging Large Language Models (LLMs, like ChatGPT) to help us scale existing Cloverleaf content to cover exponentially more scenarios, personality combinations, and even combinations of people working together. Specifically, we are exploring leveraging AI to improve the relevancy of daily coaching you receive each day, to provide more interactive and conversational coaching to help you more deeply understand the coaching tips you receive, and much more. In the spirit of transparency and proactivity, we will continuously update our customers on any technological advancements of our product and any implications they might have around data privacy and security.

At Cloverleaf, we are acutely aware that AI is at the core of technological advancement AND sometimes the speed of all this innovation can seem daunting. That’s why we’re dedicated to not only embracing the latest advancements but also making them accessible and understandable for you. When we make changes or advancements, we communicate them in plain language, so you’re never left feeling overwhelmed. We strive to clearly outline how we develop and use AI, letting you know exactly how our products function. We also address pressing concerns like data privacy, trust, ethics, and legalities head-on, openly sharing our policies to help you feel secure with our technology. At Cloverleaf, we put your comfort and understanding at the heart of our AI innovation.

The larger an enterprise organization becomes the more overhead that comes with managing the increasing number of employees. AI provides a unique ability to create personalized coaching at a scale that matches the growth pace of any organization. Our vision is not to replace humans in the coaching model, but rather to facilitate higher-quality coaching by giving managers and their team members the language and tools to have richer conversations that help them unearth the intricacies of their personalities (motivators, thinking styles, behaviors, etc.). AI enables more personalized support to meet every Cloverleaf user’s unique needs, build better relationships, and be their best selves at work.


Have questions about AI or data privacy? Contact us to learn more.