NixOS: configuration to system.
Trace a configuration change through evaluation, builds, and system activation. Examine dependencies, generations, and the limits of rollback.
How Nix worksEXPLANATORY VIDEOS BY TARUN AJJARAPU
I need to touch grass...
Operating systems, low-level code, and ML infrastructure. Detailed explanations grounded in source code and execution.
ML systems. Infrastructure. Agents.
The channel hasn’t launched yet. Planned topics are listed below.
PLANNED TOPICS
Topics for upcoming technical breakdowns.
Trace a configuration change through evaluation, builds, and system activation. Examine dependencies, generations, and the limits of rollback.
How Nix worksFollow the shell, loader, and process initialization before execution reaches main.
Benchmark a concurrent program as the worker count increases. Examine cache lines, synchronization, and scaling limits.
PLANNEDTrace a request through an inference server: compute, memory movement, scheduling, and time spent waiting.
PLANNEDBACKGROUND
I work on infrastructure and agents at Judgment Labs. Previously, I built log delivery infrastructure at Databricks and worked on payments at Stripe.
This channel covers systems programming, operating systems, and ML infrastructure.
Experience & projects