Reliability

• Keeping the Model Awake: Building a Self-Healing ML Inference Platform · 2023-02-14

  A field report on taming production machine learning inference with proactive healing, adaptive scaling, and human empathy.

• Timeouts, Retries, and Idempotency Keys: A Practical Guide · 2022-09-08

  Make your distributed calls safe under partial failure. How to budget timeouts, avoid retry storms, and use idempotency keys without shooting yourself in the foot.

• Stochastic Processes for Computer Science: Poisson, Brownian Motion, Queueing and Reliability · 2022-02-12

  A rigorous treatment of continuous-time stochastic processes—Poisson processes, CTMCs, Brownian motion with the reflection principle—and their applications in queueing theory, reliability engineering, and network performance.