Modern workflow orchestration for data and ML engineers. Get the code: https://xitter.cf/t.co/QSWkoBOIBt Chat in Slack: https://xitter.cf/t.co/XpBioqgBty
Joined November 2017
- Tweets2.2K
- Following152
- Followers6.9K
- Likes2.6K
Prefect retweeted
Can Kilo Code actually use your API?
We ran Kilo Code (Nemotron 3 Ultra) on 124 live production APIs.
On the 84 APIs every agent completed, it has the highest average score of all seven agents.
Top: @DeepgramAI and @OpenRouter at 95, then @PrefectIO, @FireworksAI_HQ and @ElevenLabs tied at 93
Full evals: devtoolarena.com/kilo
@
ALT devtoolarena. can kilo code use your api. we benchmark across 150+ developer tools. powered by lightsage agent-led growth infrastructure.
first time all together in person. we were more than a little excited, so we went to New Orleans.
here's what went down…
1. all the @dagster folks met all the prefect folks face to face for the first time
2. we took over a french quarter balcony and gator tour boat
3. at our unconference, people tried coke mixed with coconut cream. controversial, yes. delicious, also yes?
we swapped stories, made big plans, and went home as one happy team. we'll share those big plans soon. stay tuned.
thank you, NOLA.
.@kepler_ai_hq replaced its homegrown orchestration stack with prefect in four weeks.
• pipeline changes in 1–2 days, down from 1–2 weeks
• 20m+ documents processed
• close to zero orchestration maintenance, down from fixes about twice a week
read the case study → prefect.io/blog/how-kepler-c…
read the case study → prefect.io/blog/how-kepler-c…
your agent says it followed every step. your agent also says a lot of things.
@jlowin and wayne grigsby will be talking about agent governance and how to enforce access rules and required steps as agents take on recurring business processes.
october 8 · 8:15 a.m. pt / 11:15 a.m. et
30-minute talk + 15-minute q&a
save your seat in the link below. ⤵️
Prefect retweeted
This just in: @sydneyrunkle sat down with @jlowin and @zzstoatzz from @PrefectIO to talk about the new MCP spec + FastMCP, Jev, evals, and more!
Give it a listen!
youtube.com/watch?v=AuAwd5TF…
Prefect retweeted
had a great convo w/ @jlowin and @zzstoatzz on lots of hot topics in the agent space:
* the new MCP spec, and how to use it w/ FastMCP and @LangChain!
* evals (and why they're hard)
* decision models like jev
fastmcp pod is up! in this episode, @jlowin and @zzstoatzz sat down with @sydneyrunkle from @LangChain to talk mcp, building agents, and how ai is changing open source.
they get into 👇
• why langchain chose fastmcp to connect agents to mcp servers
• keeping open source collaborative when maintainers would rather use their own coding agents to write the code
• how to tell whether an mcp tool actually helps an agent
• why sydney thinks agent development could look mostly like designing evals a year from now
• experiments with jev for tool selection, model routing, and evals
watch the full episode ↓
fastmcp pod is up! in this episode, @jlowin and @zzstoatzz sat down with @sydneyrunkle from @LangChain to talk mcp, building agents, and how ai is changing open source.
they get into 👇
• why langchain chose fastmcp to connect agents to mcp servers
• keeping open source collaborative when maintainers would rather use their own coding agents to write the code
• how to tell whether an mcp tool actually helps an agent
• why sydney thinks agent development could look mostly like designing evals a year from now
• experiments with jev for tool selection, model routing, and evals
watch the full episode ↓
still recovering from 4 days at dbt summit? same.
thanks to everyone who spent time with us in las vegas. whether you stopped by to chat about one of our three products, created a custom hoodie, or dug into agentic workflows with our team, we’re glad we got to see you.
the dbt community made our booth the place to be. we’ll see you next year.
when you least expect it, JLo comes at you lightning fast with a new python library 🐍
Jev is amazing: fast, simple, and powerful.
So I built a Python DSL to match.
Introducing ✨✅ vibecheck.
Four functions that drop decision models right into your code: check, classify, label, and score.
github.com/jlowin/vibecheck
