Welcome to the New User Experience
Everything you need to rapidly prototype with AI, in a single platform. Build, collaborate, and integrate NUX Workbooks into your app.
What's a NUX Workbook?
Combine modular hand-offs between models, code and data to quickly build powerful applications.
Easy Inference
- Foundation Models: Integrate multi-modal foundation models like Llama and Whisper.
- Fine-Tuned Models: Include your own task-specific models.
- Production-Grade: Versioning, testing and parameter optimization in a single interface
Glue Code
- Python & JavaScript: Execute glue code effortlessly between stages.
- Pre/Post Processing: Access variables from any stage within your custom functions.
- Easy Version Control: Git-like version control for complete traceability.
Data Sources
- Bring Your Own KNN: Already have an embedding store? Bring it using one of our plugins.
- 3rd Party APIs: Have an existing API you already use? Bring it!
- Secure Handling: Full control and secure management.
What are others building?
...and YOU can too!
Integrate
Easily integrate data sources and models for a unified workflow. Supported Sources
Build
Create and test complex stage combinations with just a few clicks. Explore Templates
Deploy
Deploy your powerful workbook via API or native SDK to wow your users. Example Outcomes
# add data sources
nux.source(["s3://url", "db://table", ...])
# create workbook
workbook_id = nux.workbook(["stage_1", "stage_2" ...])
# integrate into your app
nux.run(workbook_id, user_request)
Focus on your users, let us handle...
Stage Handoffs
Synchronization, queuing, and serialization for a smooth, error-free transition between each stage of your chain.
Version Control
Full git-like tracking for every change in model, code and search across your end-to-end pipeline for full traceability.
Inference and Scaling
Performing inference on GPUs, while handling high request volume and maintaining industry SLAs.
We also play nice with your existing stack
NUX Strengths
Faster, cheaper and more accurate. With a single line of code.
Improve Speed
By distributing the extraction and search workloads, our servers index and search faster than the competition.
Reduce Spend
Guaranteed 30% more affordable than your current file indexing and searching approach.
Increase Conversions
Through state-of-the-art Large Language Models (LLMs), we're able to understand user intentions.
Full Model Orchestration
Supporting Your AI Prototyping
Model Chaining
Select from existing ML models or bring your own.
Version Control
Git-commit like interface for every stage-level change.
BYO KNN
Connect your existing KNN data store, supported stores
Code Execution
Inject "glue code" as Python or Javascript in between every stage.
Full SDK
Run each workbook via REST API and integrate natively in your app.
Multi-Modal
Provide or return text, image, video and audio content from your workbooks.