Aman Singh on Why Arca Lets AI Agents Prompt Humans
Arca head of AI platform Aman Singh explains why building an AI-native financial firm beats selling SaaS, and why agents should prompt humans.
Using Agents for Long-Horizon Financial Work — Aman Singh, Arca
With Aman Singh, Richard MacManus · Sunday, November 1, 2026
Arca head of AI platform Aman Singh speaks with Richard MacManus at AI Engineer New York about building an AI-native wealth management firm. Singh outlines why Arca chose to operate as a full-service firm rather than a SaaS vendor, how their custom harness manages multi-day agent workflows with persistent memory, and the guardrails required to keep human experts in the loop for financial trade execution.
Key takeaways
- Arca operates as a direct wealth management firm instead of a SaaS vendor, giving the team authority to mandate operational workflows and staff feedback loops. Read more →
- Arca operates as a full-service wealth management firm instead of a software vendor, giving leadership direct authority to mandate staff feedback on every agent output. Read more →
2 articles from this episode
Arca Forces Staff to Explain Every AI Override
Aman Singh explains how Arca turns wealth manager overrides into continuous agent evaluation datasets.
More Latent Space episodes
- How Anthropic ships the #1 LLM in the World — with Katelyn Lesse and Angela JiangNov 1 · 4 articles
- Building ChatGPT for Finance — Joseph Kim, OpenAI's Head of Financial ServicesNov 1 · 6 articles
- An Eval for AI Scientists That Knows the Answer Key - Ian Quigley, Leash BioNov 1 · 4 articles
- Why Agents Love Supabase — with Paul Copplestone and Ant WilsonNov 1 · 5 articles