About the project:
The idea of the project is to adapt AI frameworks and platforms to meet the client’s needs, allowing for quick adjustment and integration.
- Turnkey “MVP” product: Fast, easy, and cheap to get running, with a baseline that can be enhanced or switched between different LLM modes without changing the user interface.
- Privacy: Ability to operate in a completely private mode with all system elements residing in a customer’s VPC, datacenter, or on edge devices for secure environments or highly confidential data.
- Customization: Capable of producing a customer-branded interface with connectivity to customer data and prompt engineering on a same-day basis, useful for rapid prototyping.
- Support for numerous different LLMs: Unlike competitors that use a single LLM, our project can support various LLMs such as Llama2 for different use cases, including fully private models for sensitive data.
- Flexibility: the project is ideal due to its flexibility, even though it may be more complex to deploy compared to special-purpose AI solutions. It is not intended to be “all things to all people.”
- Customizable UI: The user interface can be tailored with a customer’s branding, including logos, colors, and text.
- Support for various “roles”: Capable of defining and using roles that include specific LLMs, prompts, live data source connectors, and vector storage systems, allowing one server to support multiple roles.
Requirements:
- At least 5 years of extensive experience in JavaScript
- In-depth knowledge of Node.js, including server-side development
- Understanding and practical experience with LangChain
- Solid comprehension and practical knowledge of the Fastify framework
- Proven experience in developing applications using the Svelte framework.
- Proficiency in implementing real-time features using Socket.io
- Understanding of the Vite build tool
- Ability to solve complex problems and challenges in a systematic and efficient manner
- Strong communication skills for effective collaboration within the team