Introducing VALD AI

AI, the VALD Way

Artificial intelligence (AI) has become embedded in daily life for many, improving how industries collect, interpret and act on information. However, for physiotherapists, performance coaches and sport scientists, most AI products have not been designed to sufficiently prioritize accuracy, safety and efficiency in their roles.

Since the rise of publicly available large language models (LLMs) like Claude, Gemini and ChatGPT, many practitioners and organizations have begun incorporating AI into their research and analysis workflows to gain faster access to information and performance insights.

Throughout this rapid period of adoption, VALD has taken a measured approach, waiting to engage until AI could be applied in a way that meaningfully improves practitioner workflows without compromising context, reliability or decision-making quality.

…VALD has taken a measured approach [to AI], waiting…until AI could…[improve] practitioner workflows without compromising context, reliability or decision-making quality.

A judicious approach to technology integration has been a consistent theme for VALD since its founding in 2015. Just as practitioners are skeptical of relying on exact diagnoses or prescription recommendations from LLMs, VALD has taken a considered approach to AI integration. This means weighing the risks of misinformation, data security concerns and AI hallucinations against the benefits of augmenting practitioners’ capabilities.

We are proud to announce that VALD AI (Closed Beta) is here: a product deliberately designed to serve both early adopters and skeptics. VALD AI is being built to help practitioners understand their data faster and with more clarity, allowing them to make the most informed decisions possible. Rather than claiming it will revolutionize the health and performance industry, VALD AI is focused on enabling practitioners to work more efficiently with the data they already trust.

We are proud to announce that VALD AI (Closed Beta) is here…

…VALD AI is focused on enabling practitioners to work more efficiently with the data they already trust.

Reducing friction and expanding optionality in how practitioners explore, visualize and understand their own data, VALD AI can support better decision-making for the patients and athletes they work with.

The Problem VALD AI Is Meant to Solve

Advanced measurement technologies such as ForceDecks, ForceFrame, HumanTrak and the suite of VALD systems produce detailed, high-resolution data. For each patient or athlete tested, this data can include hundreds of unique pieces of information per test, all consolidated in cloud-based platforms like VALD Hub.

Overwhelmed by metrics and daunted by a steep learning curve, practitioners who are newer to performance and rehabilitation technology may increasingly rely on publicly available LLMs to help interpret and manage large volumes of testing data. However, fabricated outputs and AI hallucinations have introduced understandable concerns around AI integration in technology.

For example, a practitioner may upload patient testing data to a publicly available LLM and receive inaccurate recommendations that influence clinical decisions, while also introducing privacy and security concerns. These workflows are often inefficient, requiring repeated exporting and back-and-forth between platforms. In contrast, VALD AI allows the entire workflow to remain within one ecosystem, making it more efficient and secure.

…VALD AI allows the entire workflow to remain within one ecosystem, making it more efficient and secure.

Time, clinical understanding and the cognitive load of day-to-day practice all influence practitioner decision-making. Therefore, VALD AI is designed to simplify how practitioners explore, interpret and interact with their data, helping improve understanding, reduce cognitive load and save time in practice.

Practitioners can use VALD AI to identify meaningful patterns and insights more efficiently through in-depth data analysis and visualizations that do not currently exist in VALD Hub. This includes the ability to explore relationships between metrics, summarize longitudinal trends and identify changes that may otherwise remain hidden with standard reporting features. These functions ultimately support faster and more informed decision-making.

Practitioners can use VALD AI…to explore relationships between metrics, summarize longitudinal trends and identify changes that may otherwise remain hidden…

Doing Few Things Exceptionally Well

VALD AI has been scoped to deliver the most immediate and reliable value in its beta stage. The VALD AI closed beta program is meant to validate performance, test usability and ensure platform reliability. In its current form, VALD AI is designed to do the following:

What VALD AI does:

• Allow users to query their own data with simple, conversational prompts
• Enable slicing and aggregation of data into clear tables or charts
• Generate internal, normative-style summaries grounded in the user’s data

What VALD AI does not do:

• Create decision-making rules or provide prescriptions for practitioners
• Generate rehabilitation programs or return-to-sport recommendations
• Share your data or other users’ data for deeper analysis

These capabilities focus on reporting clarity and data visualization, two areas where faster, more efficient outputs can support more effective decisions with greater confidence.

[VALD AI will] focus on reporting clarity and data visualization, two areas where faster, more efficient outputs can support more effective decisions with greater confidence.

AI models and systems can occasionally generate variable results, especially as their scope broadens. This expanding range of applications often increases the risk of incorrect, misleading or fabricated outputs, known as hallucinations. Taking a deliberate, focused approach to AI integration helps minimize these issues early, while prioritizing accuracy, transparency and repeatability as VALD AI continues to evolve.

This systematic process will be the blueprint for VALD AI’s approach. New features and capabilities will be deployed only when quality, safety and reliability can be assured.

Importantly, VALD AI is not purely a branded wrapper around an external LLM. While it integrates established AI infrastructure, VALD has built its own workflows, logic and domain-specific strategies into the system to better align outputs with practitioner needs, data-interpretation requirements and real-world performance and rehabilitation contexts.

Safe Exploration

VALD AI operates within VALD’s existing data protection framework. This means VALD AI is protected by the same high privacy and security standards that apply to all other VALD products, as outlined on the VALD Trust site.

VALD AI provides a more secure, purpose-built environment for interrogating VALD data than publicly accessible LLMs, with guardrails around data privacy, security and isolation designed specifically for practitioner workflows. Rather than exporting sensitive data outside VALD, where privacy, ownership and monitoring cannot be controlled, practitioners can explore and analyze their data safely within VALD’s protection standards.

VALD AI in the VALD Ecosystem

Group Dashboard Monitoring Charts, Quadrant Charts, Dynamic Strength Index (DSI) Reports and Timeline View will all remain core components of VALD Hub, with more functionality and features constantly being added.

VALD AI is designed to complement these existing reporting and communication tools by extending how practitioners explore and interrogate data, while VALD Hub will continue to serve as the purpose-built platform for all VALD users.

Similarly, VALD APIs, valdr and valdrViz, will continue to evolve to support the users who fall within the “Let me do it” persona. However, VALD AI is built to match the flexibility of advanced programming options while retaining the simple, consistent VALD user interface that practitioners can intuitively navigate.

…VALD AI is built to match the flexibility of advanced programming options while retaining the simple, consistent VALD user interface…

VALD AI supports how practitioners already work with their data by reducing the friction of custom reporting. This allows practitioners to analyze data in ways that best suit their workflow and interpretation style, helping decision-making become simpler, faster and more effective.

Expanding with Purpose

Over time, VALD AI will continue to expand, but that expansion will prioritize quality and security over speed. Future iterations will be guided by practitioner needs, supported by evidence and implemented only when their value is clear and trustworthy. This approach ensures that new capabilities solve real problems rather than adding complexity or novelty for its own sake.

We want practitioners to spend more time doing what they do best – working with patients and athletes – and less time creating reports, completing documentation and performing administrative tasks. Therefore, as with all VALD solutions, VALD AI is designed to support and augment practitioner expertise by improving efficiency, enhancing data interpretation and simplifying day-to-day workflows.

Interested in VALD AI? Express your interest in the closed beta program and stay tuned for updates.