Home Artificial Intelligence Memory Leak — #17 🚀 Products 📰 Content 💼 Jobs

Memory Leak — #17 🚀 Products 📰 Content 💼 Jobs

Memory Leak — #17
🚀 Products
📰 Content
💼 Jobs

VC Astasia Myers’ perspectives on machine learning, cloud infrastructure, developer tools, open source, and security.

Coda began a waitlist for its ­­­­alpha version of Coda AI that summarize meeting notes & transcripts in a snap using GPT3. It’s stackable with Coda’s other constructing blocks like tables, controls, text, and formulas.

Incumbents are quickly adopting foundational models to reinforce existing products. We consider that there will even be a wave of generative AI native SaaS corporations that can win. SaaS corporations that don’t adopt foundation models won’t have the identical fatality rate as on-premise software corporations that didn’t move to SaaS.

Last November, Databricks announced the provision of the Security Evaluation Tool (SAT) for AWS. Recently they announced that SAT is offered for Databricks customers on Azure and GCP. SAT helps their customers harden their Databricks environments by reviewing current deployments against our security best practices. It uses a checklist that prioritizes observed deviations by severity and provides links to resources that help resolve outstanding issues. SAT may be run as a routine scan for all workspaces in your environment to assist establish continuous adherence to best practices, and health reports may be scheduled to offer continual confidence in the safety of all data, including your sensitive datasets.

A handful of knowledge security posture management corporations have sprung up over the past yr. It is especially necessary when there may be sensitive data. Many infrastructure corporations evolve to have security product lines over time like Cisco and VMware. It can be interesting to see how Databrick’s broadens its security solutions over time.

Source: https://www.databricks.com/blog/2023/02/03/announcing-multi-cloud-support-security-analysis-tool-sat.html

Ever have trouble remembering shell commands and flags for this or that? Ever wish you might just say what you would like the shell to do? Don’t worry: GitHub is constructing GitHub Copilot assistance right into your terminal.

In January 2023, Microsoft Chief Executive Satya Nadella said that greater than 1 million people had used Copilot to this point. We’ve discussed the rise of terminal technologies prior to now here. It’s smart and unsurprising the GitHub wants to focus on the 60 million developers who use the terminal with Copilot, which has been wildly successful.

Jordan Tigani, co-founder and CEO of MotherDuck, make the case that the era of Big Data is over. It had a great run, but now we will stop worrying about data size and deal with how we’re going to make use of it to make higher decisions. A few years ago he did an evaluation of BigQuery queries, customers spending greater than $1000 / yr. 90% of queries processed lower than 100 MB of knowledge.

MotherDuck helps teams use DuckDB, an in-process SQL OLAP database management system. The thesis is that almost all corporations don’t have to scale out data processing instances because they really aren’t computing a number of data. It’s another perspective to Spark. DuckDB has grow to be a highly regarded open source project, and in line with PyPI stats was download ~900K last month.

Source: https://motherduck.com/blog/big-data-is-dead/

LangChain announced an integration with Chroma, a vector store and embeddings database designed from the ground-up to make it easy to construct AI applications with embeddings.

ML practitioners use OpenAI’s Embedding API to generate language embeddings, after which index those embeddings in vector databases for fast and scalable vector search. This can be a powerful and customary combination for constructing semantic search, question-answering, threat-detection, and other applications that depend on NLP and search over a big corpus of text data. LangChain provides foundational model orchestration to enable these pipelines, and Chroma is a latest vector database so the combination makes a ton of sense.

Seventy-one percent of all organizations run databases and caches in Kubernetes, representing a 48% year-on-year increase. Along with messaging systems (36% growth), organizations were increasingly using databases and caches to persist application workload states.

Kubernetes continues to grow in popularity, while sentiment is mixed regarding the system in line with our Twitter survey of 306 people. One cache solution that may run with Kubernetes is Dragonfly.

⭐️Claypot — Founding Engineer (Infra)

⭐️Grit — Design Engineer

⭐️ Speakeasy — Founding UX Lead



Please enter your comment!
Please enter your name here