Build GenAI applications 10x faster with Hasura

With Hasura Notebook get generative AI querying over your structured private data sources via GraphQL and REST APIs

Build GenAI applications 10x faster with Hasura

Dev challenges

Complexities of building AI apps

Building data APIs for your AI apps can present a number of challenges related to data consistency, security, scalability, and performance.

Time consuming

Building data APIs for multiple data sources with authentication and authorization, like role-based access control is time consuming and not optimized for performance, scalabililty, or reliability.

Security is not easy

Querying vector datastores in a secure manner is challenging. Vector datastores are not connected to application data model, making it difficult to query vectors for specific entities or applying access permissions.

Integration is complex

Vector datastores with additional metadata require data to be synced between application database and vector datastore. There is a high risk of the vector datastore going out of sync leading to stale results.

Simplify the process

Transform the way you build AI apps with Hasura

For features that help build APIs and backend systems for your AI applications, look no further than Hasura’s platform.

Apply authentication and role-based access controls that provide security support on your data APIs out of the box.

Combine multiple data stores like Application DB and Vector DB to access data through a single unified query.

Query Vector DB with filter condition and permission settings from your relational data model.

Set up business logic like auto-vectorization on your App DB to keep your Vector DB in sync.

Set up custom LLM logic as a Hasura Action to expose a secure API.

Transform the way you build AI apps with Hasura

More features at a glance

APIs over VectorDBs

Hasura enables you to integrate with vector databases of your choice using a custom data connector. You can play with data connectors for Weaviate and Qdrant.

Authorization with predicate pushdowns

Query your Vector DBs with filters and permission rules from your relational data model. Authorization at the query generation layer helps in faster performance and secure data access.

Knowledge API with multiple data sources

Build a unified knowledge API by combining multiple data stores like your application DB, Vector DBs, and external APIs and join across them.

Prototype faster with Jupyter Notebook

Set up custom LLM logic as Hasura Action to expose a secure API. Leverage Hasura’s Notebook data connector to build and test custom logic in Python. Leverage models from Huggingface, OpenAI to build production- ready apps.

Automatically vectorize with events

Set up business logic like auto-vectorization on your application database like PostgreSQL or SQL Server to keep your Vector DB in sync. Instantly query your structured data that is getting vectorized with Hasura Event Triggers.

Sumesh Meppadath

Sr. Engineer, Data Science, Product Engineering at GalaxEye

“Now, with the revolutionary Vector DB integration, Hasura has democratized AI's power through GraphQL, empowering our engineers to build and operate at scale with unprecedented agility, to unleash unmatched geospatial AI capabilities.

Speed, innovation, and transformative power - that's what Hasura brings to the table.”

Sample apps for building with Generative AI and Hasura

HR app powered by GenAI

In this course, explore using a stack comprised of Hasura, a Vector DB service, and Next.js paired with Tailwind to quickly create an engaging and helpful tool to support HR managers.

This tool uses leverage the power of near-text queries from vectorized data and even allow querying of LLMs to return human-like responses on data.

Contextual product search powered by OpenAI

Hasura Notebook is a remote Jupyter Notebook with ready-to-run examples and Jupyter Kernel Gateway baked-in to turn your code cells into endpoints.

Check out a demo of contextual product search powered by LLM in an e-commerce app where the search matches the user query with the product description.

GenAI powered SRE Assistant

Imagine a solution that could provide SREs with a dedicated assistant capable of seamlessly handling all information that they receive from various sources. The SRE assistant is a conversational bot supercharged by Large Language Models (LLMs) designed to effectively manage the wide variety of inputs and challenges encountered in this high-stakes environment.

Ready to start your AI journey with Hasura?

Discover Hasura Notebooks, a tool designed to facilitate the swift prototyping of cutting-edge GenAI applications.