Three Data Point Thursday

Three Data Point Thursday is a Substack dedicated to enhancing business intelligence through data and AI. It explores the strategic implementation of data teams, AI advancements, data analytics, synthetic data, and open-source contributions to building data-driven companies. The newsletter emphasizes practical approaches for leveraging data for business value, innovation, and efficiency.

Data Strategy Artificial Intelligence Data Analytics Business Intelligence Open Source in Data Synthetic Data Data Engineering Machine Learning Data-Driven Decision Making Community Building in Tech

The hottest Substack posts of Three Data Point Thursday

And their main takeaways

Last Mile of Analytics, Encrypted Spark, Barr Moses; ThDPTh #71

0 implied HN points • 09 Jun 22
  1. The last mile of analytics is crucial for startups but may not be the main bottleneck for all companies.
  2. Encrypted Spark allows for computations on encrypted datasets without significant speed issues.
  3. Data observability, inspired by software engineering, helps determine system health based on outputs.

NFTs -> Data, Product Thinking, Nbdev; ThDPTh #52

0 implied HN points • 06 Jan 22
  1. The post is about NFTs and related concepts like Data, Product Thinking, and Nbdev.
  2. The post is by Sven Balnojan, shared on January 6, 2022.
  3. For more information, visit www.thdpth.com.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

🐰 Dbt; Matt Turck; Apache Airflow; ThDPTh #40 🐰

0 implied HN points • 07 Oct 21
  1. Matt Turck provided an extensive landscape of the data space, showing significant growth in the past few years.
  2. Dbt's Event Logging package helps keep track of model runs and provides valuable metrics for data analysis.
  3. HomeToGo's use of Apache Airflow showcases the tool's prominence in data orchestration, emphasizing developer experience.

Hex gets exciting! Malloy; Autonomous Systems; ThDPTh #43

0 implied HN points • 28 Oct 21
  1. Hex notebooks are worth a try for storytelling and building small apps.
  2. Hex has a strong data vision for the future of the market.
  3. The classification of levels in autonomous driving can be a useful framework for designing autonomous systems.

🐰 #35 Quick and Dirty Data Platform Guide, Dbt what’s the hype? TimescaleDB on COSS business models; ThDPTh #35 🐰

0 implied HN points • 02 Sep 21
  1. Data platform building involves steps like ingestion, storage, analytics, and more with specific tools in each stage.
  2. Debate around Dbt's future importance and the evolving role of data lakes and SQL in the data ecosystem.
  3. Commercial open-source companies like TimescaleDB emphasize the importance of broad adoption and credibility before monetization.

🐰 #23 Revive Dead Dashboards, terminusDB, DX; ThDPTh #23 🐰

0 implied HN points • 10 Jun 21
  1. Revive your dashboards by creating a single source of truth and replacing old reports with curated ones.
  2. Consider trying out terminusDB for a graph document store with versioning and a simplistic approach.
  3. Focus on improving developer experiences for data professionals, making tasks like tracking data easier.

LakeFS, Boundary-layer, SQLPad; ThDPTh #6

0 implied HN points • 11 Feb 21
  1. LakeFS provides versioning and branching for data lakes
  2. Declarative DAG tools like Boundary-layer help simplify data pipelines
  3. SQLPad offers easy SQL editing with versioned connections

Why you need to ditch dbt for SQLMesh today

0 implied HN points • 15 Jun 23
🕹 Technology Data Engineering AI Data Analytics Product Development Community Engagement
  1. Building products with LLMs is challenging and requires addressing multiple issues.
  2. PandasAI offers AI-powered features for data analysis, focusing on integrating LLMs smartly into products.
  3. Consider switching to SQLMesh from dbt, especially if you are a data engineer or data scientist needing a more developer-focused analytics tool.

Dark Data; Thoughtful Friday #27

0 implied HN points • 17 Mar 23
🕹 Technology Data Analytics Information
  1. Dark data is information collected but not utilized, similar to dark matter in the universe.
  2. There are 6 categories of data, including what is used, not used but should be, and should be collected but isn't.
  3. Having unique data, especially dark data, can provide a competitive advantage and is valuable for a company.

Zingg, Data Meshes, Dagster; ThDPTh #53

0 implied HN points • 13 Jan 22
  1. Dagster focuses on data assets, not just pipelines.
  2. Transitioning to a data mesh may not be suitable for every company at the moment.
  3. Start-ups can benefit from implementing a decentralized data culture from the beginning.

Airbyte vs Meltano, OS Book, Kestra; ThDPTh #58

0 implied HN points • 17 Feb 22
  1. New data orchestrator called Kestra is mature and scalable
  2. Feedback indicates Airbyte and Meltano may not be ideal for mature data engineering teams yet
  3. Nadia Eghbal's Open Source book provides insightful concepts for those in the open-source space

Web3 Leaders vs Data Leaders

0 implied HN points • 18 Mar 22
  1. The post compares Web3 Leaders with Data Leaders.
  2. The author is Sven Balnojan.
  3. The article was published on March 18, 2022.

🐰 #32 Please don’t become the next AltaVista, TikToks Algorithm, Snowflakes Difference; ThDPTh #32 🐰

0 implied HN points • 12 Aug 21
  1. TikTok's algorithm success is linked to how long users watch videos. It's easy to reverse engineer and could focus more on long-term wins.
  2. Snowflake differs from other platforms by truly decoupling storage from compute. This gives it a unique selling point in the market.
  3. In the data industry, open-source solutions dominate, leading to a winner-takes-all market dynamic similar to search engines.

Kensu, Maxime on DE, Pricing is hard; ThDPTh #63

0 implied HN points • 24 Mar 22
  1. Pricing is challenging in the data space and can impact alignment.
  2. Maxime is a respected figure in the industry, sharing valuable insights on data engineering trends.
  3. Kensu.io offers a data observability toolkit worth exploring.

NordicData Architecture, Functional Data Engineering, 3 Trees; ThDPTh #69

0 implied HN points • 12 May 22
  1. Functional Data Engineering emphasizes reproducibility and efficiency in data processing.
  2. Nordic Data Architectures prioritize tools with large communities for efficient data processing.
  3. Calculating website carbon emissions can raise awareness and promote environmental responsibility.

Thoughtful Friday #16: Why you should DAG twice

0 implied HN points • 26 Aug 22
  1. DAGs in data systems can have downsides.
  2. Data systems benefit from fault-tolerant communication.
  3. Avoid bad practices in data workflows, focus on output.

Crypto Collapse, Great Dbt Models, Gartner D&A Summit; ThDPTh #80

0 implied HN points • 24 Nov 22
  1. Data concierges over data plumbers is important - deliver the right data, not a lot of data
  2. Components of well-written dbt models include modularity, readability, and speed
  3. Even well-written code decays over time; refactor often to maintain quality

🐰 #36 Tabular Icebergs, Firebolt & Data Meshes; ThDPTh #36 🐰

0 implied HN points • 09 Sep 21
  1. Tabular received Series A funding to work on Iceberg, an analytical table format.
  2. Firebolt aims to be faster than Snowflake with unique data handling approaches.
  3. Gloo.us implemented a Kafka-based data mesh to decentralize data ownership and track data versions.