In today’s world, data drives commerce. Companies like Google and Amazon organize vast amounts of information about products and services, making it easy for consumers to search and buy. However, businesses often have little control over how their products or services are presented.


The inventor of the web, Tim Berners-Lee, had an idea to address this. It’s called the Semantic Web. The Semantic Web encourages businesses to organize product information into structured data using standards like RDF (Resource Description Framework) and OWL (Web Ontology Language). These formats make it easier for machines to understand information, enabling users to search and buy products. However, this approach demanded significant manual effort—organizing products and services—without clear, immediate benefits. As a result, it was not widely adopted.


We aim to change this by making the Semantic Web relevant in the emerging agentic world—a future where intelligent agents handle daily tasks for users. In this post-Google search era, digital assistants (or agents) in consumer apps will go beyond displaying affiliate links or advertisements. Instead, they will actively assist users in making decisions and completing actions.


For example, imagine a fitness app helping you purchase healthy food. Instead of offering generic ads or links to e-commerce websites, it directly curates a shopping list tailored to your nutritional goals, finds the best deals at local stores, and even places the order on your behalf. This agent-centric functionality will depend on structured, machine-readable data that businesses provide about their products and services. In this new paradigm, our mission is to:


Generate metadata about businesses worldwide, enabling them to turn customer conversations into actionable data.


Start with Forms

The Semantic Web failed because it offered no immediate benefit to business owners. What if we start with forms that all businesses already use? Forms are the foundation of metadata, creating a structure for capturing data from users. However, creating and managing forms can be time-consuming and technical for many businesses. What if we create forms automatically by pulling metadata directly from their printed forms, existing websites, or offline brochures? Even better, let’s generate them as Google Forms that most business users are already familiar with.


For example, Formesign allows psychiatrists to upload their intake forms, which were previously printed and signed by patients. Once uploaded, Formesign automatically converts these forms into Google Forms with eSignature and HIPAA compliance, ready for patients to fill online. This makes metadata generation seamless and effortless for hospitals.

Convert Conversations


Once these forms are created, hospitals can ask their patients to fill them out. However, a patient suffering from depression might not be in the mood to fill a long intake form. Instead, they may already have this information in the referral letter from their family doctor, which they can easily send via email. What if we could extract the information from that email attachment and prefill the form, so the patient only needs to review and submit? This is what we call Semantic AI.


Promptrepo transforms such unstructured communications into structured data, guided by metadata from forms. This approach allows hospitals and businesses to benefit from both structured data and personalized interaction with their customers.

Connect Customers

This doesn’t have to stop with Formesign. We can extend this to Neartail and Formfacade as well. For instance, in Neartail, we can help businesses create their order forms automatically by pulling metadata directly from their existing websites or offline brochures. Similarly, a customer might prefer to send a WhatsApp message to repeat last week’s order rather than fill out a form again. Here too, we can transform these unstructured WhatsApp messages into structured data.


Now, businesses can experience the immediate benefit of structured data by tracking delivery dates and sales performance through reports. But how does it help their customers as promised by the Semantic Web? Our search engine for healthy food offers a glimpse of the future. It uses Promptrepo to extract nutrient information from web pages and ranks food based on NutriScore. This exemplifies how agents can focus on niches such as healthy eating (Neartail’s focus) and expand into end-to-end wellness for customers from food to fitness to healthcare (Formesign’s focus).

Summary

The Semantic Web was ahead of its time, proposing a new way to structure and connect data. However, its lack of immediate returns made it impractical in an era dominated by centralized platforms like Google. It required businesses to rewrite content designed for humans into formats for machines. Now, with advancements in large language models (LLMs), computers can understand human-written content, and the Semantic Web’s potential can finally be realized.


Our belief is that once the generative AI hype settles down, businesses will switch to Semantic AI as their primary use case for AI—moving beyond chat interfaces that mimic ChatGPT to agents that work with structured data, fulfilling the original promise of the Semantic Web. In this future, we aim to empower businesses to thrive by:
  1. Effortlessly capturing metadata using tools that create forms automatically.
  2. Seamlessly transforming unstructured conversations into structured data, guided by metadata.
By bridging the gap between the Semantic Web's promise and the agentic world's demands, we are building a future where businesses retain control over their data, streamline processes, and turn every interaction personalized with their customers — all without adding unnecessary work for their teams.