Your Finance Agent's Bottleneck Is You — Ramana Siddanth Emani, Auditoria AI
Ramana Siddanth Emani from Auditoria AI argues that developers themselves are often the bottleneck when scaling production agents for finance—not the models or hardware.
He demonstrates how to use coding agents to automate your own developer loop: agents can run in parallel on separate code branches, pull tasks from GitHub, Jira, and QA reports, write tests, run build processes, and report results. With a human as final verifier, you can ship much more code while gradually becoming less necessary in the process. At Auditoria, they use the same pattern for financial agents that communicate with each other and reconcile data.
Open on YouTube →
Vibekollen prepared this summary with AI from the original publication. The content belongs to AI Engineer.
