Frameworks
7 posts
Trust Is the Real Interface: what makes an AI root cause analysis checkable
An AI root cause analysis earns trust when you can check it: findings tied to evidence, confidence stated with its gaps, and overrides that change the answer.
The Automation Boundary: when remediation should run on its own, and when it should ask
When should automated remediation run without a person? Suggest, orchestrate and autonomous modes, and three tests: reversibility, determinism, blast radius.
Determinism, Probability, and the Limits of Observability: what the residual says about a fault
Every fault model is some version of y = mx + c + ε. What the residual's shape says about a fault, where chaos caps prediction, and why ε never reaches zero.
The hardware intelligence gap: BMC, SMART and PCIe data that nothing joins (Solution Landscape 4/4)
BMCs, SMART and PCIe counters already report on every server. Why the readings stay in separate tools and vendor dialects, and what joining them takes.
Maintenance as compounding capital: paying down hardware and configuration debt (Solution Landscape 3/4)
Why maintenance is where firmware and configuration debt gets paid, a playbook for cycles that leave the fleet more current, and what gets better with age.
GPU fleets are the new mainframes: why one failed node stalls a training job (Solution Landscape 2/4)
In synchronous training, one failed GPU node stops the whole job. What Meta's Llama 3 run shows about failure rates, and how to run GPU nodes as non-fungible.
Peer comparison is the baseline: what cohorts catch, and the faults they miss (Solution Landscape 1/4)
Why comparing each node with its peers catches degradation that thresholds miss, the cohort-wide faults it cannot see, and the four comparisons that cover them.