News

[Interview] CLEVI: Building the Future with AI Technology

CLEVI is a company showcasing a range of AI solutions, demonstrating its technological capabilities across fields including coding automation, physical AI, drone-based situational awareness, digital twins, and training data based on teleoperation and synchronisation. Through its generative AI model ‘Ivy’, it has enabled the adoption of high-performance, low-cost AI…

CLEVI is a company showcasing a range of AI solutions, demonstrating its technological capabilities across fields including coding automation, physical AI, drone-based situational awareness, digital twins, and training data based on teleoperation and synchronisation. Through its generative AI model ‘Ivy’, it has enabled the adoption of high-performance, low-cost AI, while also securing trust in its security through registration requiring minimal personal information and support for on-premises installation. Its goal is to provide a platform that enables users to implement solutions using nothing but their imagination, without specialised knowledge.

Body image

CLEVI CEO Lee Hwan-ho / Photo: Robot Technology

Solutions based on foundation models

CLEVI is an AI specialist company that has developed a foundation model with 1.4 trillion parameters, possessing technological capabilities that compare favourably with global big tech companies such as OpenAI. CLEVI CEO Lee Hwan-ho said, “Under the slogan ‘Once Fiction Now Reality’, our goal is to create the foundation that turns imagination and ideas into reality.” The key is to provide a platform that enables users to turn their imagination into reality without specialised knowledge.

CLEVI provides solutions across various fields based on its proprietary foundation model. A representative example is its advanced Coding Agent, which goes beyond simple code generation to provide automation functions that build websites from start to finish and deploy them to servers according to the user’s requests. CEO Lee Hwan-ho said, “It is technology suited to Vibe Coding, and many developers were impressed by its performance during the demonstration.”

Body image

A visitor is trying CLEVI’s Coding Agent firsthand. / Photo: Robot Technology

CLEVI’s AI technology is also making its mark in robotics and drones. In the field of Physical AI, when a person demonstrates a specific action approximately 50 times, the platform uses reinforcement learning to enable a robot to act autonomously. Previously, thousands or tens of thousands of repetitions were required for learning, but CLEVI has reduced this to 30–50 demonstrations, without direct developer intervention. In addition, by generating virtual data using its in-house image editing AI, CLEVI provides an efficient learning environment without the inconvenience of collecting real-world data.

In the drone field, specific situations such as disasters and accidents are identified, and the situation is analysed using AI predictions and camera vision. This enables the system to determine potential incidents and response methods, allowing for rapid and efficient crisis response.

Digital Twins and Teleoperation Technology

CLEVI’s technological capabilities also stand out in the digital twin field. By building virtual environments, tests that are difficult or impossible to conduct in real-world environments can be performed. Based on this, data learned through Teleoperation and synchronisation was applied and augmented, enabling demonstrations of humanoid robots. A Large Language Model (LLM) was deployed as on-device AI to enable the robot to assess situations and act like a person. Using this technology, CLEVI developed a solution that deploys robots instead of people in hazardous environments such as radioactive areas. Ultimately, the goal is to develop robots that can autonomously assess and act on commands they have not been trained on.

Main image

Industrial AI EXPO CLEVI booth overview / Photo. Robot technology

CLEVI has also addressed the issue of AI development and operation costs. Launched in March 2025, the generative AI model ‘Ivy’ is a high-performance, low-cost model that reduced the burden of AI adoption for businesses, with development costs of approximately KRW 1.5 billion and average monthly operating costs of KRW 10 million. Ivy provides powerful features including real-time learning, multi-data processing, a code interpreter and expert model generation, and recorded high scores in benchmark tests against Open AI, DeepSeek and others.

Meanwhile, CLEVI has also established reliability in terms of security. To reduce concerns about personal information leaks associated with the increased use of AI assistants, the Ivy app allows users to sign up with only the minimum personal information. Public institutions and businesses can also build security systems suited to their environments through on-premises installation.

CEO Lee Hwan-ho emphasised, “Ultimately, CLEVI’s goal is to create a platform,” adding, “We will provide a platform that allows people to create their own solutions through imagination, even without specialised knowledge.”

Reporter Im Seung-hwan

Monthly Robot Technology [email protected] <Copyright © Korea General Technology Co., Ltd. Unauthorised reproduction and redistribution prohibited>

Source: Robot Technology (Reporter Im Seung-hwan) | https://robotzine.co.kr/entry/276139

Back to newsroom
CLEVI

Language and region

Machine-translated languages are marked. Availability follows the published site bundle.

136 languages

Recommended

1

East Asia

7

Southeast Asia

11

South Asia

18

Central Asia

5

Middle East and the Caucasus

10

Western and Southern Europe

16

Britain and Ireland

4

Northern Europe and the Baltics

10

Central Europe and the Balkans

14

Eastern Europe

5

East Africa and the Horn

8

West and Central Africa

9

Southern Africa

8

The Americas

5

The Pacific

5
[Interview] CLEVI: Building the Future with AI Technology — CLEVI