Raindrop AI co-founder Ben Hylak launches Raindrop 2.0, introducing self-healing agents to autonomously detect and triage agent workflow failures
Custom-trained models flag issues like looping build failures
Many users praised Raindrop AI's 2.0 launch of self-healing agents plus the new monitoring site and Workshop for impressive quality and real improvements, while a few suspected paid promotion or found the video distracting.
Most Activity
The future belongs to designers who build.
@raindrop_ai dropping a major release today and the freshest of sites. Bravo @benhylak and team!
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
i look fwd to every @raindrop_ai launch bc it means more easter eggs for me :D
oh and i guess the new agents are pretty slick ;)
website was designed/built entirely in-house
you have to make your own landing pages, because that's when you discover everything wrong with your product
@benhylak congrats! video explains @raindrop_ai extremely well.
also, whoever designed / built your website deserves a raise 🤯
made with❤️
try it below and let me know what you think :)
The future belongs to designers who build.
@raindrop_ai dropping a major release today and the freshest of sites. Bravo @benhylak and team!
Beautiful night @raindrop_ai 🤍
The http://raindrop.ai website is *beautiful*, and love the product motivation.
Full transparency for agent successes (and failures) – and automatically applying quick fixes, once an issue is flagged – is much needed. 🥹 🌧️
Congrats on the release, @benhylak and team! 🙌
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
try raindrop 2.0 now: http://raindrop.ai
with raindrop 2.0, we also re-built trace support from the ground up. it's lightning fast, and very beautiful.
this completely replaces the need for dedicated LLM trace stores.
As you can probably tell, I'm a big fan of what the Raindrop team is building. They put genuine care and craft into a product that I'm a happy DAU of
If you're thinking of using them / curious about use cases beyond what I've mentioned in the video DM me!
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
evals don’t have to be complicated. we made them just three steps: detect, fix, verify.
your agent fails in production, raindrop catches it, your coding agent fixes it, and the failure becomes an eval so it never happens again.
raindrop 2.0 follows what we call an “agent-centric design” philosophy. agents and developers are equal-class citizens.
every convo you start with the raindrop agent can be handed off to devin/codex/claude.
everything you can do in the app can be done through an agent.
raindrop 2.0 follows what we call an “agent-centric design” philosophy. agents and developers are equal-class citizens.
every convo you start with the raindrop agent can be handed off to devin/codex/claude.
everything you can do in the app can be done through an agent.
i sat down with @banisgh, an ai engineer at vercel, and an early-tester of @raindrop_ai 2.0.
As agents take on more responsibility, reliability becomes critical.
@raindrop_ai is working to prevent that.
They just released Raindrop 2.0 to help build fixes out of failures. Their “self-healing” agent loop evaluates multiple angles of the problem, with the goal of preventing it from happening again.
We first invested in Raindrop in 2025, and co-founders @alexisgauba, @benhylak, and @snarkyzk have continued to innovate with agents in production.
As more businesses turn to agents, tech like this that can help improve quality and durability becomes indispensable in the stack.
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
i sat down with @banisgh, an ai engineer at vercel, and an early-tester of @raindrop_ai 2.0.
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
@benhylak congrats! video explains @raindrop_ai extremely well.
also, whoever designed / built your website deserves a raise 🤯
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
try it yourself: http://raindrop.ai
website was designed/built entirely in-house
you have to make your own landing pages, because that's when you discover everything wrong with your product
with raindrop 2.0, we also re-built trace support from the ground up. it's lightning fast, and very beautiful.
this completely replaces the need for dedicated LLM trace stores.
raindrop 2.0 also brings support for the features that enterprises demand.
our fortune 100 customers are already using SSO/SAML/audit-logs and more.
we are also announcing a beta-program for our VPC offering (AWS only). contact me/sales if interested.
http://www.raindrop.ai
made with❤️
try it below and let me know what you think :)
raindrop 2.0 also brings support for the features that enterprises demand.
our fortune 100 customers are already using SSO/SAML/audit-logs and more.
we are also announcing a beta-program for our VPC offering (AWS only). contact me/sales if interested.
evals don’t have to be complicated. we made them just three steps: detect, fix, verify.
your agent fails in production, raindrop catches it, your coding agent fixes it, and the failure becomes an eval so it never happens again.
raindrop is so good and shipping so fast. really the only player in this category that we trust
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?
it's incredibly hard to find agent quality issues in the wild, and even harder to triage them. @raindrop_ai has built an amazing product that solves both
huge kudos to @benhylak, alexis, zubin, and team, and very cool to see claude code in @warpdotdev in the wild!!
today, we’re excited to announce raindrop 2.0: self-healing agents.
we now train custom models that autonomously detect hidden issues with your agent.
i could tell you all about it, but wouldn’t you rather hear it from someone… else?