Python is the preferred programming language in artificial intelligence; however, most organizations are unable to integrate Python developers into legacy data infrastructure. This means the organization is unable to reap the benefits of AI. However, there has been a dearth of open-source Python libraries for AI workflows designed for developers.
The Berlin-based startup DLT (short for data load tool) may have the answer. It is developing the open-source Python library that it claims is designed for this new AI onslaught.
Read also: The value of fintech has decreased: What’s the next step for them?
The startup claims that its library will integrate into existing workflows, such as Google Colab notebooks, AWS Lambda functions, Airflow DAGs, GPT-4-assisted documents, and GPT-4 development environments, where data loading previously did not occur.
Dig Ventures, founded by Ross Mason, creator of the Mule Project and founder of MuleSoft (MULE: NYSE), has provided $1.5 million in pre-seed funding to the startup. AI and enterprise founders from companies such as Huggingface, Instana, Miro, and Matillion are participating in the round.
CEO Matthaus Krzykowski emailed me: “The majority of the GPT-4 applications featured in the media are demoware. Users who test them soon abandon them. Other AI tools where VCs have recently invested heavily (vector databases and frameworks) face many of the same obstacles.”
According to him, dlt now has a growing community of Python developers and is “deployed in production in several scale-up tech companies,” including Harness, a software delivery company based in San Francisco that we previously covered.
Read also: I attempted to purchase a post
Alexander Butler, Senior Data Engineer at Harness, stated in a statement, “Using dlt has altered our data operations.” It has… increased the speed of our DataOps team: we spend less time on the EL (extract and load) and more time on the T, while still being able to thoroughly customize our extractors as business requirements change.”
The Cambrian explosion of Python open-source tools that have become so accessible to practitioners, according to Julien Chaumond, CTO/Co-founder at Huggingface and angel investor at dltHub, has enabled the current machine learning revolution. As an easy-to-use Python library, diet is the first instrument this new group of people can employ.”
Krzykowski acknowledges that the ventures compete with Meltano, Stitch Data, Airbyte, and, to a lesser extent, Fivetran.
Read also: CoinFund raises $158M for a web3 and cryptocurrency fund.
However, he states that “from a broader perspective, we operate in the same space as data warehouse companies like Snowflake, Databricks, and Microsoft Fabric that also want to build AI-powered enterprise solutions.”
