Databricks APJ Hackathon Intelligent Apps User Guide

Databricks APJ Hackathon Intelligent Apps

 Overview

An APJ-wide virtual hackathon for teams to build intelligent applications on Databricks that combine data, AI, analytics, and automation to address real-world business problems. Challenge tracks include Social Impact (Open Data) and Business Impact (Your Data).

Program Details & Platform Elements (Source-Grounded)

Quick Start

1. Choose a track: Track 1 (Social Impact using open/public data) or Track 2 (Business Impact using your organization’s data/system tables).
2. Select your workspace approach: use an existing company/enterprise workspace, or apply for a Databricks Express Account (eligible teams receive $700 USD per team; credits begin distribution on 27 April and thereafter within 72 hours of application).
3. Apply for the Express Account by submitting the application form; only one team representative needs to submit.
4. Build an Intelligent Application using Databricks technologies (for example, Lakebase, Genie Spaces, Databricks Apps, Agent Bricks).
5. Prepare submission materials including: project/team name, participant names, country of submission, and a use case description.
6. Create a functional demonstration video (maximum 5 minutes), set it to public, and submit it via a public link (YouTube/Vimeo). Ensure the video includes: team intro + problem statement, solution architecture walkthrough, and demonstration of the working solution.
7. Submit by the submission deadline: 22nd May 2026.

APJ Hackathon

Welcome to the Databricks Intelligent Apps Hackathon, in collaboration with AWS!This APJ-wide virtual hackathon invites app developers, data and AI engineers, analysts, and business users to team up and build intelligent applications on Databricks that solve real-world business problems. This year’s challenge is simple: go beyond dashboards and create apps that combine data, AI, analytics, and automation into experiences people can use every day. Using Databricks Apps, Genie, Lakebase, and Agent Bricks, teams will design and prototype intelligent solutions that deliver insights, automate workflows, and create new user experiences.

Track 1
Social Impact (Open Data)

• Use public/open data (e.g. national open data portals) to tackle a social challenge such as climate resilience, public health, inequality, or energy e?ciency.

Track 2
Business Impact (Your Data)

• Use your organisations data or system tables to solve a real business problem – for example, cost optimisation, customer experience, risk, operations, or enterprise processes.
• Demonstrate how intelligent apps can transform traditional business processes into adaptive self-improving systems.

Express Workspace vs Customer Workspace

You can either:

• Use your existing workspace for your company/enterprise
• Use an Express Workspace with free credits

What’s the difference:

• Express workspaces are created as stand alone and come with free credit to complete your Hackathon Project
• Company workspaces are already part of your business and may have access to your desired datasets (particularly relevant for track 2)

Choose whatever fits your use case best

How to Apply for a Databricks Express Account
Eligible teams can receive $700 USD in Databricks Express account credits per team to help you start building quickly. If you prefer, you are also welcome to use your organization’s existing workspace. Credits will begin to be distributed from 27 April, and then on an ongoing basis within 72 hours of application for eligible teams. Apply for Express Account via this url https://forms.gle/xEospVHgbnLBApsa9 Please note that only one team representative needs to submit the application form

Submission Requirements

Teams must submit:

• Project/Team name
• Participant names
• Country of submission
• Use case description.
• A functional demonstration video (maximum 5 mins), which should include:
  • Introduction of your team and problem statement
  • Walk through of solution architecture
  • Demonstration of the working solution
  • Important: The video must be set to public and submitted via a public link (YouTube/Vimeo).

The project must incorporate Databricks technologies to build an intelligent application use case. This may include Lakebase, Genie Spaces, Databricks Apps, Agent Bricks, etc. Participants can use a dataset of their choice, leveraging open-source data sources and datasets available on the Databricks Marketplace

Hackathon Timeline

• Submission Deadline 22nd May 2026

Lakebase
Fully-managed Postgres for intelligent applications

• Applications have evolved
  • Applications have become stateless, serverless, and intelligent
• Databases haven’t
  • Databases remain monolithic and impossible to move

Fully-managed, serverless Postgres

• Separates compute and storage
• Great Developer Experience Branching & snapshots
• Lower TCO Autoscaling, elasticity, serverless
• Integrated Platform Simplify your stack, with centralized governance and no ETL

Data Intelligence Platform

Branching and development

Treat your database like code

• Lakebase Branching  Instant, zero-copy isolation lets teams safely build and test without touching production and live workloads.
• Git-style workflow for data
  Bring CI/CD to your database for feature branches and rapid experimentation.
• Ephemeral environments
  Create and decommission dev/test branches quickly with no storage duplication.

Recovery and backups
Precise restoration, at scale

• Scheduled snapshots
  Create instant database backups on a schedule.
• Rapid, precise recovery Restore to an exact moment within a managed window (up to 35 days) for quick incident recovery.
• Safe validation
  Use an isolated, zero-copy branch before applying changes to reduce the risk of bad restores to production.

Autoscaling
Fast performance and smart savings, with no manual tuning

• Dynamic compute sizingCompute automatically scales based on workload demand, for high performance even during spikes.
• Configurable
  Define minimum and maximum compute size to keep costs under control.
• No downtime
  Scaling operations do not interrupt operations.

Fast startup and scale to zero
Ready when you need it, without paying for idle capacity

• Only pay for active use
  Pay only when your database is active. Idle instances drop to zero to keep costs low.
• Automatic scale to zero
  Unused databases automatically suspend after a configurable idle period with no manual intervention.
• Instant readiness
  Sub-second start times mean responsive apps and fast setup for dev/testing.

Horizontal read scalability
High concurrency without operational complexity

• Read isolation
  Dedicated read-only compute serves queries without impacting write workloads.
• Independent scaling
  Add or remove read query capacity without changing write compute.
• Shared data layer
  All compute accesses a single underlying data set, with no data duplication.

Built into the Data Intelligence Platform
Avoid separate stacks and proprietary governance models.

• First-class platform compute
  Execute operational workloads on the same platform as analytics and AI.
• Share a common data foundation Operational data is immediately available without copying or ETL pipelines.
• Built-in governance and security
  Apply consistent access control, auditing, and policies across all data.

What can you do with Lakebase?

Serve low latency and high concurrencyPersonalised recommendations, customer segmentationFeature store

Build AI and traditional appsOrder processingInteractive workflow sign-offState for an agent

Analyze data in the lakehouseOrder history (analytics) Chatbot history (training data)

Integrated with Lakehouse
Easy to use

Fully-Managed Data Synchronization

• Sync from Lakehouse table – Postgres
• Sync from Postgres – Lakehouse table
• Snapshot, Triggered, Continuous
• Multiple table syncs in a pipeline

AI Platform: Online Feature Store

• High-performance feature lookup
• Feature store for real-time serving, with offline training data in Lakehouse

Integrated with Unity Catalog

• Register Postgres database as a catalogue
• governed by UC
• Lakehouse Federation

Query Editor and Schema Browser

• Native Postgres query editor
• Browse Postgres schema

Postgres Extensions

• We support an ever growing list of extensions, which are documented on our website
• The list of available extensions can also be retrieved from Postgres directly using SELECT * FROM pg_available_extensions ORDER BY name;
• Installing an extension is as easy as CREATE EXTENSION <extension_name>;
• Amongst the popular extensions: pg_stat_statements (query metrics), pgcrypto, pg_prewarm (cache prewarming), neon, postgis (geospatial), pg_graphql (GraphQL support), pg_vector