Why AI Infrastructure POCs Stall Before Production | Rob Hirschfeld, RackN | TFiR
Rob Hirschfeld, CEO at RackN, explains why hand-built AI clusters never reach production and how Digital ...
How AI Is Making Scrapers, Scalpers, and Bots Harder to Stop | Steve Winterfeld, Akamai | TFiR
Akamai Advisory CISO Steve Winterfeld explains how AI is accelerating scraping and scalping attacks, why the friend-frenemy-enemy ...
How to Decide Where AI Inference Should Run: Edge vs. Centralized | Ari Weil, Akamai | TFiR
Ari Weil of Akamai breaks down which AI inference workloads require edge distribution, which do not, and ...
Which JDK Versions Still Receive Security Patches and From Whom | Simon Ritter, Azul | TFiR
Simon Ritter, Deputy CTO at Azul, explains how security-only monthly patches eliminate regression risk, which LTS versions ...
How to Keep Data Consistent Across MCP, SDK, and CLI AI Agents | Mario Moscatiello, Airbyte | TFiR
Mario Moscatiello of Airbyte explains why MCP, SDK, and CLI interfaces will coexist in enterprises, why consistent ...
Why AI Inference Needs Different Infrastructure Than Training | Dr. Robert Blumofe, Akamai | TFiR
Dr. Robert Blumofe of Akamai explains why AI training and inference require completely different infrastructure architectures, why ...
What to Demand From AI Infrastructure Before Scaling Inference | Jon Alexander, Akamai | TFiR
Jon Alexander of Akamai outlines what enterprises must evaluate before large-scale AI inference deployments: portability, elasticity, storage ...
How to Manage Data Services Across Multiple AWS Accounts | Julian Fischer, anynines | TFiR
Julian Fischer, CEO of anynines, explains why enterprises cannot centralize data services in a single AWS account ...
Edge AI Inference: The Last-Mile Problem Centralized Clouds Cannot Solve | Ari Weil, Akamai | TFiR
Ari Weil of Akamai explains why centralized GPU clusters cannot meet the latency demands of fraud detection, ...
How AI Inference Hardware Differs From Training Hardware | Rob Hirschfeld, RackN | TFiR
Rob Hirschfeld of RackN explains why context window management, not GPU compute, is the primary bottleneck in ...