Jul 6, 2026 · 58 min
AI startups target the manual processes holding back asset management
$100T is managed by “human duct tape” | E2308
This episode bridges the future of AI-driven fintech with essential historical lessons on SaaS pricing, security compliance, and platform dependency.
- 1Hanover Park is building an AI-native ERP to automate manual data entry and modernize legacy fund operations.
- 2Figma's early success relied on bottom-up growth, transparent billing, and navigating evolving security compliance standards.
- 3The rise and fall of Mahalo highlights the extreme vulnerability of building a business dependent on Google's search algorithms.
Don't miss
Lon Harris shares insider details on working at Mahalo and how Google's algorithm changes decimated the search engine's traffic.
The brief
The $100 trillion asset management industry still relies on manual data entry, basic spreadsheets, and human middlemen. Hanover Park co-founder Chris Cote joins the show to explain how his startup is building an AI-native ERP to modernize legacy fund operations.
Hanover Park plans to replace outdated systems by leveraging advanced AI models to rapidly ingest and clean complex financial data. This shift aims to transition CPAs from manual data entry workers into high-value, strategic advisors for asset managers.
The episode also features a March 2020 flashback interview with Figma co-founder Dylan Field. The hosts analyze Figma's early bottom-up go-to-market strategy, transparent billing models, and how security standards like SOC 2 compliance have evolved.
In a nostalgic turn, co-host Lon Harris shares his firsthand experience working at Mahalo.com, Jason Calacanis's early human-curated search engine, illustrating how sudden Google algorithm updates permanently disrupted their business model.
Whether rebuilding fund administration with AI or surviving platform shifts, the tech landscape proves that long-term success requires owning your data pipelines and building transparent, trust-based relationships with customers.
What was said on this episode
24 statements · 15 positive · 8 negative · 1 neutral
Legacy services businesses run the backbone of $100 trillion in global assets.
“Businesses that are running the entire backbone of these hundred trillion of global assets”
Listen at 2:17
Fund administration is a large underserved market with low technology penetration.
“massive untapped legacy services market with very low tech penetration”
Listen at 2:37
Traditional fund administrators leave funds dependent on human intermediaries for their own data.
“you end up in this like crazy situation where you're stuck with a bunch of human middlemen that, that are holding your own data hostage”
Listen at 3:27
Fund data is now 100 times more valuable because it underpins investment decisions.
“your data is a hundred times more valuable because this is like the backbone for every investing decision you've ever made”
Listen at 4:06
AI agents with memory can preserve fund-specific accounting knowledge when employees leave.
“building agents with memory on top is actually the way you solve that problem”
Listen at 5:27
Hanover Park’s long-horizon agents can clean financial data at scale.
“we built these Long Horizon agents to do financial data cleaning at scale”
Listen at 7:34
Hanover Park is pursuing one-click migration of fund data.
“We say, I call it the one click migration future”
Listen at 10:41
Hanover Park migrated a venture fund and launched its LPs six days later.
“We took that data, we got that data. We launched their limited partners six days later”
Listen at 11:55
One-click migration became feasible roughly three to six months ago with Opus 4.
“that was only possible probably three to six months ago with like, Opus 4”
Listen at 13:13
AI performance gaps in fund operations are driven more by context than intelligence.
“the gap is less of a intelligence gap and more of a context gap”
Listen at 13:54
Human CPAs remain necessary to review AI outputs and handle fund-accounting edge cases.
“we need to have incredible CPAs that are crushing it, that are tier one, reviewing associated outputs and handling edge cases”
Listen at 15:54
CPA advisory on unusual fund transactions is more valuable than routine journal-entry checking.
“that is high value advisory work, not. Did you book this journal entry correctly?”
Listen at 17:37
Automating routine tasks can provide real-time fund data and shift CPAs toward advisory work.
“if we're doing, if we're automating those lower value tasks and we're delivering real time data versus delayed quarters and quarters, because that's when humans are doing it, then we can be the consigliere”
Listen at 17:48
Hanover Park intends to preserve customer data integrity and not share customer data.
“we are laser focused on data integrity, not sharing your data with others”
Listen at 18:33
Traditional fund administration pricing contains many hidden, opaque fees.
“there is a lot of random hidden opaque fees that live in the market today”
Listen at 20:10
Figma can spread within organizations without requiring broad initial approval.
“your organization are able to adopt it and they're able to spread it without having to be to like necessarily get a lot of buy in from others around them”
Listen at 30:19
SOC 2 processes help cloud-hosted companies improve security.
“it's basically just a process to make sure that you're able to be as secure as possible even though you're hosted in the cloud”
Listen at 31:39
Active-user pricing incentivizes desirable customer behavior.
“It incentivizes the right behaviors”
Listen at 37:45
Google’s search changes pushed competitors down and replaced their results with answer boxes.
“go back in time to them pushing Yelp down, putting ehow, Mahalo, everybody else out of business or moving them down the page and ankling them and then replacing them with the one box”
Listen at 47:00
Google controlling AI and search could make it an arbiter of truth and speech.
“if Google ends up owning the AI market as well as the historical search market, then I think they become essentially the arbiter not only of truth but of speech”
Listen at 50:50
Early restrictive COVID measures could stabilize conditions and prevent hospital overload.
“we'll see, you know, sort of like the stabilization, the fact that you're talking about and hospitals won't be overloaded”
Listen at 53:23
Relaxing COVID restrictions could produce a second wave and renewed hospital overload.
“I think we'll see a second wave effect where there still is the virus out there, and we'll start to see it spread again, and then hospitals will be overloaded then”
Listen at 53:36
COVID restrictions in the San Francisco Bay Area could last substantially longer.
“I think it could be a lot longer potentially”
Listen at 53:53
COVID-19 dramatically accelerated the preexisting shift toward remote work.
“Covid just massively, like lit a rocket under that revolution”
Listen at 56:53
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.