Alexander Gallego, Redpanda Data | AI Infrastructure Silicon Valley - Executive Series
summary
Claims that infrastructure must move from batch to realtime and that deploying models next to data — not shipping data to foundation models — is the enterprise path. Describes Redpanda Data as an end-to-end streaming immutable log powering banks, Akamai and gaming firms; cites Llama 3 17B fine-tuned to run on 16 GB GPUs and open runtimes like VLLM/Ollama; highlights Apache Iceberg as the open table format and S3 tiered storage to cut clusters (example: a 384-node deployment to 24), calling this model-to-data and immutable log architecture the operational pattern.
AI-written summary · about
