https://alberto.codes/ https://alberto.codes/about https://alberto.codes/experience https://alberto.codes/projects https://alberto.codes/publications https://alberto.codes/blog https://alberto.codes/blog/2026-10-01-gemma-read-a-love-note-as-a-scam 2026-10-01 https://alberto.codes/blog/2026-09-29-gemma-was-sure-the-total-matched-it-had-not-seen-the-receipt 2026-09-29 https://alberto.codes/blog/2026-09-27-jev-looked-underconfident-two-labels-were-the-reason 2026-09-27 https://alberto.codes/blog/2026-09-08-the-scan-lost-one-letter-and-crowned-a-derivative 2026-09-08 https://alberto.codes/blog/2026-09-06-serve-gemma-4-31b-on-a-24-gib-card-with-the-context-it-was-packed-for 2026-09-06 https://alberto.codes/blog/2026-09-05-the-parent-finally-has-a-verb 2026-09-05 https://alberto.codes/blog/2026-09-04-i-cut-the-last-sauce-off-the-file 2026-09-04 https://alberto.codes/blog/2026-09-03-the-book-spells-it-out-at-line-1437 2026-09-03 https://alberto.codes/blog/2026-09-02-googles-4-bit-gemma-already-fit-my-card 2026-09-04 https://alberto.codes/blog/2026-09-01-i-added-a-second-copy-of-the-same-book 2026-09-01 https://alberto.codes/blog/2026-08-22-the-2-bit-label-was-4-5-bits-inside 2026-08-22 https://alberto.codes/blog/2026-08-21-marrow-sauce-finally-has-a-parent 2026-08-21 https://alberto.codes/blog/2026-08-19-there-is-no-model-in-this-parser 2026-08-19 https://alberto.codes/blog/2026-08-15-fit-a-model-to-the-gpu-you-actually-have 2026-08-15 https://alberto.codes/blog/2026-08-15-a-different-ceiling-is-a-different-recipe 2026-08-15 https://alberto.codes/blog/2026-08-11-i-couldnt-tell-my-quantized-model-from-the-baseline 2026-08-11 https://alberto.codes/blog/2026-06-02-two-patterns-that-make-prod-db-changes-boring 2026-06-02 https://alberto.codes/blog/2026-03-31-from-one-model-to-seven-making-turboquant-model-portable 2026-09-06 https://alberto.codes/blog/2026-03-28-serve-compressed-vlm-inference-from-a-container 2026-03-28 https://alberto.codes/blog/2026-03-27-paper-to-pypi-in-72-hours-building-the-first-turboquant-vllm-plugin 2026-09-06 https://alberto.codes/blog/2026-03-26-i-ran-turboquant-on-a-vision-model-the-first-output-was-garbage 2026-03-26 https://alberto.codes/blog/2026-03-23-i-asked-an-ai-to-explain-boto3-then-i-fixed-the-docstrings 2026-03-23 https://alberto.codes/blog/2026-03-22-when-docstrings-lie-your-ai-tools-pay-the-price 2026-03-22 https://alberto.codes/blog/2026-03-06-encrypt-adk-sessions-in-five-minutes 2026-03-06 https://alberto.codes/blog/2026-03-05-stop-writing-ai-agent-prompts-by-hand 2026-03-05 https://alberto.codes/blog/2026-03-04-give-your-ai-agent-a-docstring-quality-tool 2026-03-04 https://alberto.codes/blog/2026-03-01-your-ai-agents-memories-arent-encrypted 2026-03-01 https://alberto.codes/blog/2026-02-25-your-ai-reads-your-docstrings 2026-02-25 https://alberto.codes/blog/2026-02-05-entity-resolution-is-recipe-matching 2026-02-05 https://alberto.codes/blog/2026-02-04-lineage-ids-in-multimodal-ai-pipelines 2026-02-04 https://alberto.codes/blog/2026-02-02-task-queues-idempotency-and-ai-pipelines 2026-02-02 https://alberto.codes/blog/2026-02-01-why-reflex-for-a-python-engineers-portfolio 2026-02-01 https://alberto.codes/blog/series/measuring-model-confidence 2026-10-01 https://alberto.codes/blog/series/saucier 2026-09-08 https://alberto.codes/blog/series/vramfit 2026-09-06 https://alberto.codes/blog/series/turboquant-on-vision-models 2026-03-31 https://alberto.codes/blog/series/docvet-and-docstring-quality 2026-03-23 https://alberto.codes/blog/series/encrypted-adk-sessions 2026-03-06 https://alberto.codes/blog/series/ai-pipeline-plumbing 2026-02-04 https://alberto.codes/contact