Episode 2: Master Scaling AI Coding Agents: Cut Costs with Best-of-N with @SambaNovaSystems

Your coding agent isn't burning your budget in the planning step — it's burning it in the executor, where dozens of edits, test-fix cycles, and parallel candidate solutions rack up inference calls fast. Join Kwasi Ankomah, Lead AI Architect at SambaNova Systems, for Part 2 of the SambaNova Sponsored Webinar Series, where he breaks down how parallel execution, best-of-N selection, and disaggregated serving turn unpredictable agent costs into a controllable line item. Kwasi brings 15 years across financial services, consulting, government, and tech startups to a conversation that goes deep on the engineering choices behind agentic AI in production. You'll learn: → Why the executor — not the planner — is where coding agents actually spend time and money → How fanning work across parallel executors changes the economics of running more candidate solutions → What best-of-N selection is, and how test suites pick winning candidates from parallel attempts → Why fast, affordable inference at scale is a prerequisite for test-time compute to pay off → How SambaNova's disaggregated serving architecture affects throughput, latency, and utilization tradeoffs → How to evaluate your own agent infrastructure for where executor costs are really coming from Register now to save your spot.