Mezmo is the intelligence layer for production AI that makes autonomous operations fast, efficient, and safe.
Joined August 2015
- Tweets 2,777
- Following 759
- Followers 1,733
- Likes 2,706
1,202 Photos and videos
Pinned Tweet
May 13
Introducing Builder in the Loop: the humans behind AURA, our open-source agentic harness for production ops.
First up: @hank_talks
His take? “The agent said so” is not an incident review.
Mr. Roboto approves 🤖
Read it: mezmo.com/blog/builder-in-th…
#opensource #AURA
1
1
1
99
Jun 10
Built for SREs who've seen the demo. Tired of the gap between that and prod.
Bring your laptop. Leave with a runbook.
Join us at PlatformCon in NYC!
#ruleprod #opensource
🔗platformcon.com/nyc-session/…
1
24
Jun 4
Meet Henry Andrews (@hank_talks), one of the builders behind AURA.
His philosophy: production software needs accountability. Agents should, too.
"The agent said so" is not an incident review.
🤖 Read more: mezmo.com/blog/builder-in-th…
23
Jun 2
Your SRE agent is not underperforming.
It is over-employed.
Investigate the incident. Pull telemetry. Check infra. Update the runbook. Open the PR.
One agent doing all of that? Cute.
Why agentic SRE needs orchestration.
🔗 mezmo.com/blog/why-sre-agent…
#AURA #opensource #AgenticAI
30
May 27
Your SRE agent should not have the memory of a goldfish 🤖
Our second Builder in the Loop feature spotlights Eric Lake on runbooks, investigation memory, and why starting from zero every incident is a terrible strategy.
Read more
🔗 mezmo.com/blog/builder-in-th…
#SRE #opensource
1
182
May 26
AI agents don’t invent bad decisions out of nowhere.
They inherit them.
🔸 Messy telemetry.
🔸 Missing ownership.
🔸 Stale runbooks.
🔸 Broken context.
Then they package it up with confidence.
If your agents touch prod, this one’s for you.
🔗mezmo.com/blog/when-your-age…
#SRE
1
18
May 21
Your SRE agent does not need its 147th tool. It needs boundaries.
Real incident response is not one job. Investigation, context, decisions, docs, execution. One agent doing all of it is how things get weird.
🔗 mezmo.com/blog/why-sre-agent…
#opensource
1
1
31
May 19
The latest Gartner Analyst Take makes it clear: context engineering is foundational to agentic AI success.
For SRE teams, that means one thing: AI agents are only as reliable as the telemetry operational context they can reason over.
Our take:
🔗 mezmo.com/blog/when-your-age…
#SRE
1
1
30
May 14
Production AI checklist:
✖ Dump every log into the context window
✖ Hope for the best
✖ Call it innovation
Actual production AI runs on context engineering, memory, governance, and earned autonomy.
More than vibes.
🔗 mezmo.com/blog/the-journey-t…
1
33
May 8
Missed the webinar? We got you.
Watch Andre Elizondo break down how teams are moving from AI experiments to trusted, production-ready workflows with AURA.
Open source. Better context. Less chaos.
Watch now →mezmo.com/resources/the-jour…
1
36
May 5
Exciting to see AURA recognized as critical infrastructure for the AI era.
As featured in AI Journal, AURA is helping teams bring structure, transparency, and trust to AI in production.
Because better context = better decisions.
mezmo.com/newsroom/why-open-…
#AI #opensource #SRE
1
71
May 5
Getting an AI agent to work in a demo is easy.
Getting it to behave in production is where things fall apart.
We’re breaking down how to fix it with AURA.
Open source. Less boilerplate. Better context.
Join us TODAY • 1 PM EST
platformengineering.org/even…
1
1
40
Apr 30
Getting an AI agent to work in a demo is easy.
Running it in production with real telemetry, cost control, and accuracy? Not so much.
AURA makes it practical.
Open source. Config-driven. Built to turn experiments into repeatable workflows.
Register → ow.ly/XZJ150YSZz7
1
80
Apr 28
ICYMI: Observability 360 called Mezmo a “pipeline powerhouse” and gave AURA a nod 😌
“There’s a lot to like… running agentic workflows over filtered, enriched telemetry.”
That’s AURA. Open source. No walled gardens.
Peek the repo: github.com/mezmo/aura
#Observability360
30
Apr 28
If your runbooks only exist in someone’s head, you’ve got a scaling problem.
AURA turns incidents into draft runbooks so every response makes the next one faster.
That’s how teams build real memory.
Dive into the full breakdown → mezmo.com/blog/the-runbook-p…
2
42
Apr 23
AI is moving fast. Making it reliable in production is the real unlock.
That’s where AURA comes in.
An open system of context for agents that can reason, explain, and integrate safely into real workflows.
aijourn.com/why-open-source-…
#AI #opensource
1
1
71
Apr 15
Grateful to TFiR for the conversation with Andre Elizondo on what makes agents reliable in the wild.
We cover why agents fail quietly, how context shapes outcomes, and how AURA helps agents plan, act, and self-correct.
Check it out 👉youtu.be/DiVAJGKhEfE?si=OtmS…
#SRE #OpenSource
30
Apr 14
The problem isn’t writing runbooks. It’s finding the time.
AURA turns real incident workflows into structured drafts, grounded in your systems, not guesswork.
From investigation → institutional memory.
Read how it works 👉 mezmo.com/blog/the-runbook-p…
32
Apr 8
“Just give the agent more data” isn’t a strategy.
Andre Elizondo shares what actually works, from context engineering to memory and orchestration.
If you're building real AI systems, this one's worth your time.
🎥 Watch: youtu.be/vr6OA8ietsc?si=LB7N…
#AI #SRE #opensource
58
Apr 2
AI agents aren’t failing because of models.
They’re failing because of infrastructure.
That’s why we open-sourced AURA.
Read Henry Andrews’ take → mezmo.com/blog/why-we-open-s…
#AI #opensource #aurafarming
1
41