This is the work I keep coming back to.
Before the job titles, there was a bench at IIT Roorkee. I built a grid of LEDs from a bare board, then a pair of robot legs, then a glove, and then I spent a year with a radio signal that was mostly noise. I wanted to see what the thing in front of me was doing. I still want exactly that.
So I stay near the running system. OpenTelemetry on the days a group of people can agree what to call things. eBPF on the days the application has to be left alone. A collector carries what they emit, and a store on the far side has to answer someone who is actually worried, sometimes at terabytes a day. When the subject has been a model, I have followed it onto GPUs, asking the same thing. What is it doing, while it does it?
When I can, I sit with people who are meeting this for the first time. Google Summer of Code, Cloud Native University, and the KubeCon rooms where someone would like it shown, plainly. The experiments that need more time than a product allows are in the labs.
Latest
Introducing continuous profiling in Odigos
If you would like to get in touch, or grab a virtual coffee to talk about anything, feel free to contact me.
Feel free to reach out to me or hop on a call with me.
or write me here and I will get back to you,