dbt-databricks
cube.js
dbt-databricks | cube.js | |
---|---|---|
15 | 86 | |
180 | 17,174 | |
1.7% | 0.8% | |
9.5 | 9.9 | |
15 days ago | 2 days ago | |
Python | Rust | |
Apache License 2.0 | GNU General Public License v3.0 or later |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
dbt-databricks
-
Curious if anyone has adopted a stack to do raw data ingestion in Databricks?
Our current data infra looks a little something like this: 1. Airbyte deployed on EKS for supported data connectors. I’m using the alpha Databricks connector to load directly into Unity Catalog. 1a. S3 bucket for raw landing zone storage if we cannot directly load into Databricks Managed Tables. 2. Orchestration, storage, and transformations are in Databricks. Calling out to the Airbyte api in the EKS cluster to keep all orchestrations inside Databricks. 2a. databricks-dbt for transformations & cleaning.
-
dolly-v2-12b
dolly-v2-12bis a 12 billion parameter causal language model created by Databricks that is derived from EleutherAI’s Pythia-12b and fine-tuned on a ~15K record instruction corpus generated by Databricks employees and released under a permissive license (CC-BY-SA)
-
Any suggestions for building DBT project on DataBricks?
Read this https://github.com/databricks/dbt-databricks
- dummy
-
Clickstream data analysis with Databricks and Redpanda
Global organizations need a way to process the massive amounts of data they produce for real-time decision making. They often utilize event-streaming tools like Redpanda with stream-processing tools like Databricks for this purpose.
- Next step for my career..
-
DeWitt Clause, or Can You Benchmark %DATABASE% and Get Away With It
Databricks, a data lakehouse company founded by the creators of Apache Spark, published a blog post claiming that it set a new data warehousing performance record in 100 TB TPC-DS benchmark. It was also mentioned that Databricks was 2.7x faster and 12x better in terms of price performance compared to Snowflake.
- Would you use dbt with databricks? If so, why?
-
Welcome, DataEngHack online!
databricks
-
A Quick Start to Databricks on AWS
Go to Databricks and click the Try Databricks button. Fill in the form and Select AWS as your desired platform afterward.
cube.js
-
MQL – Client and Server to query your DB in natural language
I should have clarified. There's a large number of apps that are:
1. taking info strictly from SQL (e.g. information_schema, query history)
2. taking a user input / question
3. writing SQL to answer that question
An app like this is what I call "text-to-sql". Totally agree a better system would pull in additional documentation (which is what we're doing), but I'd no longer consider it "text-to-sql". In our case, we're not even directly writing SQL, but rather generating semantic layer queries (i.e. https://cube.dev/).
-
Show HN: Spice.ai – materialize, accelerate, and query SQL data from any source
I'm not too familiar with https://cube.dev/ - but my initial impression is they are focused more on providing APIs backed by SQL. They have a SQL API that emulates the PostgreSQL wire protocol, whereas Spice implements Arrow and Flight SQL natively. Their pre-aggregations are a similar concept to Spice's data accelerators. It also looks like they have their own query language, whereas Spice is native SQL as well.
-
Show HN: Delphi – Build customer-facing AI data apps (that work)
Hey HN!
Over the past year, my co-founder David and I have been building Delphi to let developers create amazing customer-facing AI experiences on top of their data. We're excited to share it with you.
David and I have spent our careers leading data and engineering teams. After ChatGPT got popular, we saw a rush of "chat with your data" startups launch. Most of these are "text-to-SQL" and use an LLM like GPT-4 to generate SQL queries that run directly against a data warehouse or database.
However, the general perception now is most of them make for nice demos but are hard to make work in the real world. The reason is data complexity. Even smart LLMs find it difficult to reason about messy databases with hundreds of tables, thousands of columns, and complex schemas that have been built up piece-meal for years. Text-to-SQL can be a fine dev tool for data scientists and analysts, but we've seen many organizations hesitate to deploy it to end users, who never know if the answer they get one day will be the same the next.
David and I found a better way. From our time in the data engineering world, we were familiar with a type of tool called "semantic layers." Think of them like an ORM for analytics. Basically, they sit between databases (or data warehouses) and data consumers (data viz tools like Tableau or APIs) and map real-world concepts (entities like "customers" and metrics like "sales") to database tables and calculations.
Semantic layers are often used for "embedded analytics" (e.g. when you're building customer-facing dashboards into your application) but are increasingly also used for traditional business intelligence. Cube (https://cube.dev) is a prominent example, and dbt has also recently released one. They're useful because with a semantic layer, the consumer doesn't have to think about questions like "how do we define revenue?" when running a query. They just get consistent, governed data definitions across their business.
We realized that semantic layers could be just as useful for LLMs as for humans. After all, LLMs are built on natural language, so a system that deterministically translates natural language concepts into code has obvious power when you're working with LLMs. With a semantic layer, we've found that companies can get AI to answer much more complex questions than without it.
For a year now, we've been building Delphi to do just that. We've gone through a few iterations/pivots (initially we were focused on building a Slack bot for internal analytics) and are now seeing our developer-first approach resonate. We're being used to power customer-facing fintech applications, recruiting software, and more.
How do you use Delphi? The first step is connecting your database; then, we build your semantic layer on top of it. Right now we do this manually, but we're moving more and more of it over to AI. Once that's done, we have 3 main ways of using Delphi: 1) white-labeling our AI analytics platform and providing it to your customers; 2) a streaming REST API and SDKs; and 3) React components to easily drop a "chat with your data" experience into your app.
If this is interesting to you, drop us a line at [email protected] or sign up at our website (https://delphihq.com) to get in touch. Thanks for reading! Would love to hear any thoughts and feedback.
-
Apache Superset
We use https://cube.dev/ as intermediate layer between data warehouse database and Superset (and other "terminal" apps for BI like report generators). You define your schema (metrics, dimensions, joins, calculated metrics etc) in cube and then access them by any tool that can connect to SQL db
-
Need to reduce costs - which service to use?
also check out cube.dev. they can do the semantic layer and cache it so you are not hitting Snowflake all the time.
-
Anyone with experience moving to Cube.dev + Metabase/Superset from Looker ?
We need metrics to live in source control with reviews. Metabase doesn't have a git integration for metrics, which is why we are convinced to use cube.dev as a semantic layer.
-
GigaOm Sonar Report Reviews Semantic Layer and Metric Store Vendors
https://github.com/cube-js/cube comes out very well at the end as a promising open source system, getting rather close to the bullseye. Would love to know more & hear people's experience with it.
-
Show HN: VulcanSQL – Serve high-concurrency, low-latency API from OLAP
How is this different from something like https://cube.dev/
-
Best Headless Chart Library?
Have a look to cube.js
-
Advice / Questions on Modern Data Stack
For now, I've been thinking on using self-hosted Rudderstack both for ingestion and reverse ETL, cube.dev as the abstraction later for building webapps and providing catching for the BI layer, and dbt for transformations. But I have doubts with the following elements:
What are some alternatives?
dbt-spark - dbt-spark contains all of the code enabling dbt to work with Apache Spark and Databricks
Apache Superset - Apache Superset is a Data Visualization and Data Exploration Platform [Moved to: https://github.com/apache/superset]
Neo4j - Graphs for Everyone
Elasticsearch - Free and Open, Distributed, RESTful Search Engine
Trino - Official repository of Trino, the distributed SQL query engine for big data, formerly known as PrestoSQL (https://trino.io)
Druid - Apache Druid: a high performance real-time analytics database.
TimescaleDB - An open-source time-series SQL database optimized for fast ingest and complex queries. Packaged as a PostgreSQL extension.
Redash - Make Your Company Data Driven. Connect to any data source, easily visualize, dashboard and share your data.
sql_to_ibis - A Python package that parses sql and converts it to ibis expressions
Metabase - The simplest, fastest way to get business intelligence and analytics to everyone in your company :yum:
nutter - Testing framework for Databricks notebooks
metriql - The metrics layer for your data. Join us at https://metriql.com/slack