Bionic

Run Gen AI Locally in Seconds

Try our Docker Compose installation. Ideal for running AI locally and for small pilots.

Familiar Chat Experience

• 🖥️ Intuitive Interface: Our chat interface is inspired by ChatGPT to ensure a user-friendly experience.
• 🌈 Theme Customization: The theme for Bionic is completely customizable allowing you to brand Bionic as you like.
• ⚡ Ultra Fast UI: Enjoy fast and responsive performance from our Rust based UI.
• 📜 Chat History: Effortlessly access and manage your conversation history.

AI Assistants (Retrieval Augmented Generation)

• 🤖 AI Assistants: Users can create assistants that work with their own data to enhance the AI.
• 🗨️ Share Assistants with Team Members: Generate and share assistants seamlessly between users, enhancing collaboration and communication.
• 📋 Agentic RAG Pipelines: Assistants are full scale enterprise ready Agentic RAG pipelines that can be launched in minutes.
• 📑 Any Documents: 80% of enterprise data exists in difficult-to-use formats like HTML, PDF, CSV, PNG, PPTX, and more. We support all of them.
• 💾 No Code: Configure embeddings engine and chunking algorithms all through our UI.
• 🗨️ System Prompts: Configure system prompts to get the LLM to reply in the way you want.

Teams

• 👫 Teams: Your company is made up of Teams of people and Bionic utilises this setup for maximum effect.
• 👫 Invite Team Members: Teams can self-manage in a controlled environment.
• 🙋 Manage Teams: Manage who has access to Bionic with your SSO system.
• 👬 Virtual Teams: Create teams within teams to
• 🚠 Switch Teams: Switch between teams whilst still keeping data isolated.
• 🚓 RBAC: Use your SSO system to configure which features users have access to.

Defence in Depth Security

• 👮 SAST: Static Application Security Testing - Our CI/CD pipeline runs SAST so we can identify risks before the code is built.
• 📢 Authorization RLS - We use Row Level Security in Postgres as another check to ensure data is not leaked between unauthorized users.
• 🚔 CSP: Our Content Security Policy is at the highest level and stops all manner of security threats.
• 🐳 Minimal containers: We build containers from Scratch whenever possible to limit supply chain attacks.
• ⏳ Non root containers: We run containers as non root to limit horizontal movement during an attack.
• 👮 Audit Trail: See who did what and when.
• ⏰ Postgres Roles: We run the minimum level of permissions for our postgres connections.
• 📣 SIEM integration: Integrate with your SIEM system for threat detection and investigation.
• ⌛ Resistant to timing attacks (api keys): Coming soon.
• 📭 SSO: We didn't build our own authentication but use industry leading and secure open source IAM systems.

Observability and Reporting

• 📈 Observability API: Compatible with Prometheus for measuring load and usage.
• 🤖 Dashboards: Create dashboards with Grafana for an overview of your whole system.
• 📚 Monitor Chats: All questions and responses are recording and available in the Postgres database.

Token Usage Limits and Controls

• 📈 Fairly share resources: Without token limits it's easy for your models to become overloaded.
• 🔒 Reverse Proxy: All models are protected with our reverse proxy that allows you to set limits and ensure fair usage across your users.
• 👮 Role Based: Apply token usage limits based on a users role from your IAM system.

Turn AI Assistants into APIs

• 🔐 Assistants API: Any assistant you create can easily be turned into an Open AI compatible API.
• 🔑 Key Management: Users can create API keys for assistants they have access to.
• 🔏 Throttling limits: All API keys follow the users throttling limits ensuring fair access to the models.

Manage Data Governance with GuardRails

• 📁 Batch Guardrails: Apply rules to documents uploaded by our batch data pipeline.
• 🏅 Streaming Guardrails: LLMs deliver results in streams, we can apply rules in realtime as the stream flies by.
• 👾 Prompt injection: We can guard against prompt injections attacks as well as many more.

Local or Remote Large Language Models

• 🤖 Full support for open source models running locally or in your data center.
• 🌟 Multiple Model Support: Install and manage as many models as you want.
• 👾 Easy Switch: Seamlessly switch between different chat models for diverse interactions.
• ⚙️ Many Models Conversations: Effortlessly engage with various models simultaneously, harnessing their unique strengths for optimal responses. Enhance your experience by leveraging a diverse set of models in parallel.

Role Based Access Control

• ⚠️ Configurable UI: Give users access or not to certain features based on roles you give them in your IAM system.
• 🚦 With limits: Apply token usage limits based on a users role.
• 🎫 Fully secured: Rules are applied in our server and defence in depth secured one more time with Postgres RLS.

Data Integrations

• 📤 100s of Sources: With our Airbyte integration you can batch upload data from sources such as Sharepoint, NFS, FTP, Kafka and more.
• 📥 Batching: Run upload once a day or every hour. Set the way you want.
• 📈 Real time: Capture data in real time to ensure your models are always using the latest data.
• 🚆 Manual Upload: Users have the ability to manually upload data so Agentic RAG pipelines can be setup in minutes.
• 🍟 Datasets: Data is stored in datasets and our security ensures data can't leak between users or teams.
• 📚 OCR: We can process documents using OCR to unlock even more data.

Deploy to Bare Metal or The Cloud

• 🚀 Effortless Setup: Install seamlessly using Kubernetes (k3s, Docker Desktop or the cloud) for a hassle-free experience.
• 🌟 Continuous Updates: We are committed to improving Bionic with regular updates and new features.

Try it out

follow our guide to running Bionic in production.

Configuration

Set the APP_BASE_URL environment variable to the public URL of your Bionic server. This value is used when constructing OAuth2 callback URLs.

Architecture

flowchart LR
    %% External actors
    User((User))
    IdP["External<br/>Identity Provider"]
    LLM["LLM Provider<br/>Ollama / OpenAI / Anthropic"]
    S3["S3-Compatible<br/>Object Storage"]

    %% Kubernetes boundary
    subgraph K8s["Kubernetes<br/>EKS · AKS · GKE · k3s"]
        direction LR

        %% Ingress & Auth
        Nginx["Nginx Router"]
        OAuth2["OAuth2 Proxy"]

        %% Core API
        RustServer["High Performance<br/>Rust Web Server"]

        %% Engines
        RAG["High Performance<br/>RAG Engine"]
        DocEngine["Rust Document<br/>Engine"]

        %% Database
        Postgres["Postgres<br/>Relational DB & PgVector"]
    end

    %% Request flow
    User --> Nginx --> OAuth2 --> RustServer

    %% Authentication flow
    OAuth2 --> IdP
    IdP --> OAuth2

    %% Control & inference
    RustServer --> RAG
    RustServer --> LLM
    RustServer --> Postgres

    %% RAG orchestration
    RAG --> Postgres
    RAG --> DocEngine

    %% Object storage access
    RustServer --> S3

Enterprise

For companies that need better security, user management and professional support

Talk to the founders

This covers:

• ✅ Help with integrations
• ✅ Feature Prioritization
• ✅ Custom Integrations
• ✅ LTS (Long Term Support) Versions
• ✅ Professional Support
• ✅ Custom SLAs
• ✅ Secure access with Single Sign-On
• ✅ Continuous Batching
• ✅ Data Pipelines

Support / talk with founders

• Schedule a Chat 👋
• Connect on Linked in 💭

Scales to 1000's of users.

Bionic is optimized to run on Kubernetes and provide Generative AI services for potentially 1000's of users.