Make AI Workflows 2x Faster and 6x Cheaper with Functional Python (Session for Americas & Europe) скачать в хорошем качестве

Make AI Workflows 2x Faster and 6x Cheaper with Functional Python (Session for Americas & Europe)

Functional programming for AI can make your ML pipelines faster and cheaper. Learn how persistent caching and deterministic parallelization cut runtime without buying more hardware. In this webinar, host Sharice Mayer talks with Vlad Pavlov, a hands-on AI engineering leader and open-source enthusiast, with 20+ years at startups to Microsoft & Intel. Vlad shares practical patterns for AI engineers and data teams to speed up experiments, reduce cloud spend, and reuse results across teams using pure functions, memoization, and a shared cache. 🔥 Key Highlights: – Why brains beat servers: replace compute with storage to win speed and cost – Pure functions 101: side-effect free code and referential transparency in practice – Persistent memoization: compute once, reuse forever across sessions and teammates – Deterministic parallelization: safer scaling to many workers without race conditions – Using underutilized hardware: turn idle machines into a background worker swarm – Where this shines: HPO, repeated data prep, ensemble evaluation, long jobs – Deployment tips: local folders, NFS/EFS, S3 style backends, and tradeoffs for latency 👥 Speaker, Host, and Organiser: – Vlad Pavlov – Senior Director, AI & Software Engineering at PMG:   / vlpavlov   – Sharice Mayer – ML & Data Scientist at Aerendir:   / sharice-mayer   – Dmytro Spodarets – Founder & CEO at Data Phoenix:   / spodarets   🔗 Helpful Resources: – Slides and Colab demos: https://docs.google.com/presentation/... – Pythagoras: https://pypi.org/project/pythagoras/ 👉 If you enjoyed the discussion, be sure to like the video, share your favorite insights in the comments, and subscribe to stay updated on our latest talks and demos! 🔗 Follow us for more: – LinkedIn (Dmytro):   / spodarets   – LinkedIn (Data Phoenix):   / data-phoenix   – YouTube (Events):    / dataphoenixevents   – YouTube (Dmytro):    / @dmytrospodarets   – Twitter/X: https://x.com/Data_Phoenix – Telegram: https://t.me/DataPhoenix – Facebook:   / dataphoenix.info   – Website: https://dataphoenix.info/