Mission & Vision of Qualiteg
Deepening the creativity of people around the world
We believe that creativity is a crucial element that drives the development of society and promotes individual growth.
Qualiteg will provide a comfortable world through technology that allows people to generate new ideas and enhance their creativity.
As represented by the Industrial Revolution, the greatest outcome of technological innovation in industries is "time".
The time created through technological innovations has allowed people to increase their creative time.
AGI is said to transform all industries, with an impact far surpassing that of the Industrial Revolution and the Internet Revolution.
People will not only regain their creative time but also be able to maximize their potential.
Solutions by Qualiteg
Qualiteg believes that LLM (Large Language Models) are the most crucial technology for realizing AGI.Therefore, our company provides a platform for running LLMs, along with various tools and services to accelerate many LLM projects, and consulting to launch the world's first AGI from Japan.
- Provision of LLM Platform ChatStream™ ChatStream™ is our platform for running Large Language Models (LLM).
- AI Consulting We provide business and technology consulting for the use of generative AI (Generative AI) from zero experience.
Our AI Solutions
【One-stop LLM Service】
You can select from thousands of LLMs to execute them quickly, securely, and with high quality, supporting your service launch with a short TAT.
You can use open-source LLMs, your proprietary LLMs, and also ChatGPT (GPT4 Turbo API).
【No-Code LLM App Building】
In addition to the execution platform, a variety of front-ends are also pre-set, enabling the construction of commercial-quality conversational applications with excellent UI/UX without code.
【Flexible Execution Environment】
Building fully managed LLM applications is possible not only by using our cloud environment equipped with cutting-edge GPUs but also in your environment, on-premises, GCP, AWS, HuggingFace, and other AI execution environments.
We also offer OEM supply as IoT embedded solutions or Python libraries.
Business consultants familiar with AI business and technology specialists will accompany and support your company's problem-solving.
We also support the introduction of LLM conversational chats.
Services
Check our Products and Services
LLM Platform ChatStream
We develop and provide "ChatStream", a commercial service-oriented LLM platform.
This product serves as a commercial LLM execution platform that can utilize proprietary LLMs or open-source LLMs as its engine.
(It is not a chat framework wrapped around ChatGPT or other such engines.)
Thus, it allows for the construction of powerful applications using strong open-source-based LLMs like Meta Llama2 and Mistral in a one-stop, short-TAT manner. It makes it possible to realize LLM applications that leverage the advantages of powerful and flexible open-source LLMs without being confined to the limitations of ChatGPT (GPT4 API).
Features:
・Can build full-fledged LLM applications with no-code/low-code.
・Wide support for proprietary LLMs and open-source LLMs.
・Flexible support for multi-user concurrent access and high-load environments through proprietary GPU load distribution technology.
・Customizable chat interface with high UX.
【Demo】
Please experience our demo!
ChatStream Demo
【Technical Information】(Japanese)
Please refer to the following for information on ChatStream's technology and how to build LLM chats using Python:
ChatStream Technical Information
【Consultation on Implementation & Setup】
Please contact us through our dealer network or the following contact point:
Contact Us
【Specifications】
・Python: Operates in a Python 3.11 environment
・Supported GPU: NVIDIA CUDA 11.7–equipped GPUs
・Chat UI: Standard-equipped multitask-capable web chat UI. Compatible with web and mobile. Customizable. Supports flexible changes to the conversation tree such as regeneration and re-editing.
・Multi-GPU Load Distribution: Supported
・Multi-Node Load Distribution: Supports clustering load distribution across multiple nodes
・Security: ASN filtering, IP filtering, TLS3.0, CSRF protection, and more
・Advanced Features: Multitasking, OAuth user authentication, and more
・Execution Environment: Qualiteg GPU Cloud (our GPU environment), AWS, Azure, GCP, on-premise (GPU procurement and HPC construction services available separately through our company)
【List of Verified Models】※
・meta-llama/Meta-Llama-2 series
・meta-llama/Meta-Llama-3 series
・lmsys/vicuna-13b series
・lmsys/vicuna-7b series
・mistralai/Mistral-7B series
・databricks/dolly-v2 series
・mosaicml/mpt-*-chat series
・togethercomputer/RedPajama-INCITE series
・Qwen/Qwen-14B series
・Rakuten/RakutenAI-7B series
・rinna/nekomata series
・rinna/youri series
・rinna/japanese-gpt-neox series
・tokyotech-llm/Swallow series
・elyza/ELYZA-japanese-Llama-2 series
・stabilityai/japanese-stablelm series
・matsuo-lab/weblab-10b series
・line-corporation/japanese-large-lm series
・cyberagent/calm2 series
・cyberagent/open-calm series
・llm-jp/llm-jp-13b series
・OpenAI Inc./GTP4-Turbo
・Cohere Inc./Command R+
※ A partial list of LLMs verified to work properly. Testing involves generating several turns of text to confirm correct text generation, but does not guarantee full operational functionality. System development and negotiations with model providers may be necessary during production service construction, depending on requirements and needs.
Examples of Support
AI Business Consulting
Examples of Support (Business)Support for launching new businesses using ChatGPT
Support for considering new services using ChatGPT
Market research for services and adjacent areas using ChatGPT
Developing generative AI courses for client companies
Considering new services using Stable Diffusion
Support for business transformation PoC using generative AI
Support for developing AI talent cultivation programs
Other experienced business consultants (MBA holders) will fully support your company's challenges
AI Technology Consulting
Examples of Support (Technology)Technical advisory in your company's AI domain
Support for building chat systems using open-source Large Language Models (LLM)
Support for setting up deep learning environments (GPU-equipped high-performance machines)
Support for containerization of image processing AI and proof of concept with Hugging Face Space
Support for fine-tuning open-source Large Language Models (LLM) for your business
Other, experienced experts will fully support your company's challenges in technical areas such as
Large Language Models (LLM), foundation models, generative AI orchestration, edge LLMs, domain-specific generative AI models, Transformers, Self-Attention, LangChain, MRKL (Multi-Round Knowledge Loop), Prompt engineering, self-supervised learning, SFT, RLHF, fine-tuning, Multi-GPU, multi-node LLM, Hugging Face Docker, LLM-based AI agents, Autonomous AI agents, Advanced Data Analysis (Code Interpreter), Open Interpreter, PyTorch, mathematics of deep learning
Company
Company Profile
| Name | Qualiteg Inc. |
| CEO | Tom Misawa |
| Business | Software research and development, Consulting and training, HPC sales |
| Location | 1-10-6 Kajicho, Chiyoda-ku, Tokyo |
| Established | July 2023 |
| Capital | 3,000,000 JPY |
| Main Bank | Mizuho Bank |
CEO's Background
After completing graduate school,
Joined Sony Corporation in 2003.
Involved in launching numerous services, serving as a producer, AI technology development leader, etc.
Founded Qualiteg Inc. in 2023.