Transcending the Limits of Human-Machine Interactions: 'Vercept' The AI Agent Turning Demonstrations into Automation

We are entering a new era shifting from a time when humans had to adapt to computers, to one where AI is expected to understand human context. This paradigm shift lies at the core of “The Future of Human-Machine Interactions”, a session hosted by SCB 10X in collaboration with Kiana Ehsani, CEO & Co-founder of Vercept. Vercept is redefining established norms by setting a new standard for communication between humans and AI, aiming to make these interactions more natural, intuitive, and seamless than ever before.

Radical Simplicity: A Vision of Ultimate Simplicity 

With over a decade of experience in AI and robotics at AI2 (Allen Institute for AI), Kiana Ehsani identified a pivotal shift: it is time for AI to move beyond competing for high scores on leaderboards (like ImageNet) and instead bring ‘Visual Intelligence’ out to create real-world business impact.

This led to the founding of Vercept under the vision of ‘Radical Simplicity.’ Their ambitious goal is to reduce reliance on traditional peripherals like mice, keyboards, and touchscreens. Vercept developed “Vy,” an AI Agent that acts as a bridge between "User Intent" and "Digital Systems," automatically transforming mental desires into completed tasks.

“Vercept’s goal is to make the mouse and keyboard a thing of the past, leaving only your intent, which is instantly translated into work.”

Beyond Prompt Engineering: The New “Watch-and-Repeat” Standard 

A major hurdle in current AI adoption is the need for users to master Prompt Engineering to get accurate results. Vercept flips this process with a ‘Watch-and-Repeat’ approach.

Vy’s mechanism doesn’t require users to write code or set up complex Robotic Process Automation (RPA) projects. Instead, the system allows users to simply ‘show it once.’ Vy watches the workflow on the screen, understands it, and memorizes the steps. If exceptions or errors occur, the user simply intervenes and teaches it new steps, just like training a human colleague, without needing to overhaul backend systems.

Visual Understanding 

The technical edge Kiana emphasizes is Visual Understanding. Vy is trained to understand the screen visually like human eyes rather than reading the underlying code (DOM). This allows Vy to work seamlessly across platforms, whether it’s Legacy Software with no APIs or modern web applications, instantly understanding the meaning of various buttons and icons.

Security and True Multitasking 

While the market focuses on Cloud or Browser-based systems, Vercept differentiates itself with On-Device Processing. This is a key strength for enterprise clients regarding:

• Data Privacy: Sensitive data never leaves the machine.
• Seamless Agility: If you can log in, Vy can work immediately without waiting for complex System Integration from IT.

Another highlight is Parallelization, which unlocks true Multitasking. Users can order Vy to run Background Tasks such as pulling data into Excel, sending emails, or scheduling appointments while they continue to attend online meetings or work on other tasks on the same screen without interruption.

Transforming Non-Techies into Super Users 

Vercept proves it isn’t just a tool for repetitive tasks, but a key to turning general users (Non-Tech) into Super Users. They can manage complex technical tasks, like advanced Excel formulas or scripts, simply by using voice commands combined with visual demonstrations.

However, Kiana notes that the greatest success isn’t just efficiency metrics, but breaking down Accessibility barriers. This technology enables users with physical or visual limitations to return to work they thought was impossible, reinforcing that good technology must drive both "Productivity" and "Equality."

“The most rewarding thing for me is messages from users with physical disabilities saying they can now do jobs they never thought possible. This is proof that accessibility and productivity can grow together.”

What’s Next for Vercept? 

Looking ahead, Kiana revealed that Vercept has been redeveloped to support cross-platform functionality. Windows Beta and Mac versions are now available, allowing workflows created by users to sync across all devices.

The next step is penetrating the communication tool market. Vy will act as a Smart Router, distributing tasks to the most appropriate device. For example, a user might issue a command via mobile to have a laptop at home run code, or instruct a workplace Cloud system to process large scripts, all without worrying about device limitations.

For those interested in experiencing this new way of working, follow their progress at http://vercept.com

📹 Watch the full content here: https://youtu.be/vNEX1O1hqIo?si=Za9bfZP381BL6-Ok