Steffen Hausmann

passionate about real-time analytics. field engineer @ Materialize. used to be into olives and photography.

[Harrison Chase @hwchase17]

🔎🤖LangSmith Insights Agent

Really excited to launch our first in-product agent

This agent lives inside LangSmith and combs through traces, giving you insights into:
🧑‍🤝‍🧑how users are using your agent
⁉️how your agent may be messing up
🛃{your custom insight here}

The problem we saw was that people were launching agents... and didn't know how their users were actually using them! You put a chat box in front of people, and they may ask it anything - the surface area for agents is often super wide

In addition - agents would fail silently. They could give a bad response - this wouldn't show up in error logs, but its good to know.

If you know what look for, you can set up LLM as a judge evaluators. But what if you don't? (most people don't initially)

The best way to figure this out - as @HamelHusain says - "look at your data". But LLMs are really good at looking at your data! So can they do it for you?

This is exactly what insights agent attempts to do. It's live in LangSmith today. You can read more about it here: https://t.co/fpPrHyfajr

[Richard Artoul @richardartoul]

@samlambert The thing people don’t get too is it doesn’t just like hard fail in an easy to understand way where you can just easily cycle the node. It fails in very annoying ways like suddenly getting very slow, sometimes, maybe, if you’re not looking at it, on Wednesdays

[Marta Paes @morsapaes]

Real.

[Marta Paes @morsapaes]

I'm speaking at #Current24 next week! Come around for an overview of the CDC ecosystem + a spoiler of a new standalone, low-footprint CDC service built using Rust and @ApacheArrow. 👀

https://t.co/rbMZYKKLNr

[Materialize @MaterializeInc]

Are you performing complex queries on operational data? Materialize costs 1/20th what Aurora PostgreSQL read replicas do, while achieving 100x greater throughput and 1000x lower latency. Read our full analysis -> https://t.co/ACkjWWt0Yy

[Materialize @MaterializeInc]

Materialize vs. PostgreSQL: Materialize delivers 100x greater throughput with 1000x lower latency. Materialize wins for OLTP offload. Check out the benchmarks in our new blog -> https://t.co/R8K5jpydcx

[Arjun Narayan @narayanarjun]

I've always found the conversation around databases and inconsistency challenging to distill into simple terms. This is my attempt to build some simpler intuition.
https://t.co/cV357f0Y4i

[Materialize @MaterializeInc]

Now available in Private Preview - Bulk exports to Amazon S3! Now you can export a snapshot of your data to Amazon S3 object storage for periodic backups or further processing in downstream systems. Check it out!
https://t.co/Khhw67Oyj3

[Software Huddle @SoftwareHuddle]

We're back with another episode of Software Huddle. Join us for an engaging discussion.

Operational Data Warehouse with Nikhil Benesch 🎉

Today's episode is with Nikhil Benesch, who's the co-founder and CTO at @MaterializeInc, an Operational Data Warehouse.

Materialize gets you the best of both worlds, combining the capabilities of your data warehouse with the immediacy of streaming. This fusion allows businesses to operate with data in real-time.

We discussed the data infrastructure stuff of it, how they built it, how they think about billing, how they think about cloud primitives and what they wish they had.

Watch on Youtube: https://t.co/FhMtVb377e

@nikhilbenesch | @seanfalconer | @alexbdebrie

[Steffen Hausmann @sthmmm]

@OpenFGA @MrLarrieu @MaterializeInc No worries. It was fun working on the project. And I wasn't alone, @MrLarrieu found a bug and @frankmcsherry provided helpful comments on the required transformations.

[OpenFGA @OpenFGA]

Join us for the OpenFGA community meeting this Thursday 11th, at 3PM UTC (8AM PT/11AM ET)!

Discover features like Modular Models, Spring Integration, our research for integrating OpenFGA with @MaterializeInc, how Canonical uses OpenFGA and more!

Details: https://t.co/sumxizEned

[Steffen Hausmann @sthmmm]

@MrLarrieu What? It would be SO nice!

[Steffen Hausmann @sthmmm]

My timeline is full of PartyRocket related stuff. With all that noise, I must have missed the announcement for the service that converts a draft into a proper presentation. Can someone please remind me? It's kind of urgent... #kafkaSummit

[Steffen Hausmann @sthmmm]

@shoellinger Excellent. I'll bring my laptop to the speaker dinner and we'll have figured this out in no time!

[Steffen Hausmann @sthmmm]

Nice! Bonus points for printing it out ♥️

[chitrangna bhatt @mchitrangna]

This morning read - Best practices for right sizing your Apache Kafka clusters to optimize performance and cost by @sthmmm. Steffen talks about most common resource bottlenecks and getting to a suitable cluster based on requirements. More in 🧵 #Kafka @apachekafka

[Ian Macomber @iandmacomber]

@MaterializeInc @tryramp "They said I couldn't empower an analytics engineering team to take a rules-based fraud detection approach down from 30 min batch to 3 seconds streaming, all with SQL and dbt, and I took that personally"