robw on 🦋. Sr Eng Mgr @ grafana; AI/ML. Tweets my own. He/him.
Central France
Joined December 2008
- Tweets4.6K
- Following195
- Followers169
- Likes6.6K
Pinned Tweet
Hey friends, I built this little site to help anyone managing engineers to run 1-on-1s: 1on1s.robwhelan.com/
(or to help engineers better understand what their managers are doing, or could be doing)
Feedback welcome; hope this helps!
Wish I could share my notes with past-me.
Rob Whelan retweeted
🙌 New episode of Go Time!
💚 #AI for #Observability
😎 with @yasirekinci from @grafana
🎙 hosts @jboursiquot & @matryer
🎧 gotime.fm/335
Rob Whelan retweeted
my cofounder presented a study that showed teenagers learn better if they can sleep in a little more, and the school came back with “well we can’t adjust it now, we have a contract with the bus drivers”
oh, so the school is for the benefit of the bus drivers??
much like the port is for the benefit of the longshoremen, and the polity is for the benefit of the politicians, eh?
Rob Whelan retweeted
The Observability Survey is back for its third year, and we want to hear from you! Share your thoughts on the state of observability today — it only takes five minutes and all responses will be kept anonymous. grafana.com/blog/2024/09/18/…
Rob Whelan retweeted
Dear managers: Raising productivity is not about monitoring people. It's about motivating them.
94 studies: Surveillance fails to improve performance—and increases stress, distrust, and dissatisfaction.
Tracking people is not a substitute for respecting and valuing them.
Rob Whelan retweeted
“at least I can write about this, at least I can help people with this, at least I am uniquely in a position to make the world a better place for this” has saved me innumerable times
Rob Whelan retweeted
I expect my grandchildren to relate to respiratory illnesses like I relate to scurvy: “Wow, the past was *awful* and there is a historically contingent reason why it stayed that awful long after the tech tree would have supported eliminating this problem.”
Air purifiers in Finnish daycare reduced children's illnesses by 18% in a carefully controlled study. This implies huge potential for better public health and cost savings from sick leaves, with a simple intervention against airborne infections.
yle.fi/a/74-20089670
Rob Whelan retweeted
I can’t hear a reference to Schmidt without thinking of this: surfingcomplexity.blog/2021/…
Rob Whelan retweeted
👋 Inspired by recent conversations with friends, and based on a long history of automating away every job I've ever had (from data processing to PM work):
Am sharing a few ways that I'm using Gemini 1.5 and 2M+ tokens of context in @GoogleAIStudio to automate the boring parts of DevRel and UXR!
Reminder that you can stuff quite a bit into 2M+ tokens (hours of video, years of emails, full codebases, etc.) and that, over time, we expect 2M tokens ➡️ infinity, cost ➡️ $0, latency ➡️ near instant.
Rob Whelan retweeted
A decade ago I felt like I could find anything on the Web. Now I feel like I can barely find anything. People just don't put information on web pages anymore.
Rob Whelan retweeted
Premature abstraction is WAY more prevalent and worse than premature optimization and really handcuffs code bases for years and years.
I'm oddly proud that my team did NOT build this.
The team that did (Pyroscope / Profiles) didn't even *tell us* they were using the OSS Grafana LLM plugin we built — they just found it & used it; no notes.
🔥 We dropped it like it's hot. The new Flame Graph AI in #Grafana Cloud uses #LLMs to assist with flame graph data interpretation so you can identify bottlenecks, root causes, and suggested fixes faster!
You should read about it, take a second: bit.ly/3UTJqDr
100% & agency is so easily missed by engineers imagining ourselves as the user.
We think "it should do as much as possible for me, that's ideal".
But the actual user didn't build it, doesn't trust it, & will happily accept it doing *less* if they keep control & understanding.
Tangent from an interesting thread — I've been trying to avoid the hard/soft skills phrasing, because it lets people devalue people skills, think they can't be learned, have no measurable results, etc..
Instead I've just been calling them "competencies", and naming them.
Some examples: you bring some exciting but half-formed ideas to a few other teams to get input, and they give you a flood of strong opinions that'd totally change your direction. Some want to work on it.
1) avoid sharing 🌧️
2) build the competencies to navigate this well 🌟
If we can first assume there's only one person coding in the world, on one project, then we can *finally* get some definitive answers to these debates.
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
