Dec 17, 2025 · 41m · mixergy
#2288 She had people do AI’s work
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Helen Hastings, founder and CEO of Quanta, explains how she built a next-generation AI accounting platform by first performing manual bookkeeping workflows, architecting an immutable double-entry ledger, and strictly focusing on modern tech-enabled businesses. She details her operational playbook for automating domain knowledge, overcoming customer trust barriers, and delivering real-time financial intelligence.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Andrew holds 27% of the talking time here. How this is scored →
Andrew as informed peer 4.2 Guest teaching 5.3 Guest disagreement 1.6 Andrew pushing back 2.2
speaking balance: gold is Andrew, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 26:32 Rejecting the human-heavy services trap
Helen forcefully rejects the horizontal strategy of earlier accounting startups, arguing they built human consultancies on top of QuickBooks rather than genuine automated software.
Hardest push from Andrew ▶ 29:44 Challenging homepage positioningAndrew inspects Quanta's live homepage and directly challenges Helen on where the supposed niche qualification actually exists for visiting prospects.
Biggest teaching moment ▶ 33:28 Accounting for AI API marginsHelen educates the host on modern SaaS economics, explaining how traditional bookkeepers misclassify LLM usage tokens as operating expenses rather than cost of goods sold.
Andrew holds their own ▶ 22:00 Citing history of failed accounting disruptorsAndrew demonstrates deep subject domain knowledge by listing his angel investments in inDinero, past sponsorships from Bench and Pilot, and questioning why anyone thinks QuickBooks can be unseated.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Andrew as informed peer | Guest teaching | Guest disagreement | Andrew pushing back | Why |
|---|---|---|---|---|---|---|
| Deploying LLMs for Semantic Understanding in Financial Bookkeeping | 4 | 5 | 1 | 2 | Andrew brings up a specific prior example regarding AWS versus Amazon purchases to illustrate LLM extraction. Helen explains how financial bookkeeping is fundamentally about understanding human-written memos and context rather than just crunching numbers. | |
| Uncovering Inefficiencies of Manual CSV Uploads and Plaid | 4 | 6 | 2 | 2 | Andrew shares his personal experience and annoyance with bookkeepers manually handling CSV files. Helen details the technical flaws of Plaid and QuickBooks, explaining how Quanta uses continuous redundant reconciliation to prevent ledger corruption. | |
| Insights from Affirm on the Power of Double-Entry Ledgers | 3 | 5 | 1 | 1 | Andrew prompts Helen on the origin of Quanta's core architecture. Helen details her tenure at Affirm building custom immutable double-entry ledgers piping into NetSuite and realizing all companies suffer from delayed financial visibility. | |
| The Downside of Delayed Closes in Fast-Moving Startups | 4 | 5 | 1 | 1 | Andrew highlights the pain of waiting weeks for end-of-month closes and having to query humans for basic software spend totals. Helen expands on how manual outsourcing creates multi-month backlogs that ruin decision-making speed. | |
| Pre-Selling Customers and Adopting the Hybrid Software-Services Model | 4 | 5 | 2 | 3 | Andrew pushes on whether verbal willingness to pay is sufficient validation given complex accounting edge cases. Helen explains the modern AI playbook of using hybrid human CPAs to bridge the last mile before automating. | |
| Overcoming Accounting Trust Hurdles with the QuickBooks Backup Strategy | 4 | 4 | 1 | 2 | Andrew references an interview with Zeni to address customer hesitation regarding accounting software trust. Helen explains their QuickBooks sync backup feature that lets reluctant clients keep a mirrored fallback. | |
| Evaluating Incumbents: Xero International Moat and QuickBooks Desktop | 6 | 5 | 2 | 3 | Andrew demonstrates significant industry knowledge, citing his angel investment in inDinero, sponsorships from Bench and Pilot, and past interviews with Xero. Helen explains Xero's international tax localization moat and why desktop legacy users resist cloud cruft. | |
| Avoiding the Startup Graveyard Through Strict Ideal Customer Profile | 4 | 6 | 3 | 4 | Andrew directly challenges Helen on how Quanta avoids the graveyard of failed QuickBooks competitors. Helen articulates Quanta's disciplined ICP strategy, contrasting it with competitors who scaled headcount to support horizontal retail clients on QuickBooks. | |
| Standardizing on Modern Tech Stacks and Avoiding Physical Inventory | 5 | 5 | 2 | 4 | Andrew checks Quanta's live website during the interview and points out the lack of explicit ICP filtering on the homepage. Helen acknowledges the point, clarifies their focus on software without physical inventory, and explains the complexities of equity accounting. | |
| Managing Volatile API Costs and Margins in the AI Era | 4 | 7 | 1 | 1 | Andrew suggests that the platform shift to AI has opened buyers' minds to replacing legacy systems like NetSuite. Helen educates Andrew on how volatile LLM API costs are distorting SaaS margins if bookkeepers misclassify them as OpEx instead of COGS. | |
| A Three-Step Playbook for Automating Manual Work with AI | 4 | 5 | 1 | 1 | Andrew synthesizes Helen's three-step playbook for automating manual workflows. Helen explains Prism's agentic reporting capability, demonstrating how it provides auditable evidence linking answers back to raw transaction sources. |
Statements from this episode (21)
Hastings: Accounting is largely understanding business context, not just numbers
“People think that accounting is just numbers. Actually, so much of it is just understanding what's going on in the business. And so much of that is actually human written.”
Hastings: Traditional bookkeeping fails because outsourced bookkeepers lack software context
“So one reason that traditional bookkeeping tends to be so painful is that it's often outsourced to people that are maybe overseas and they, they're great, but they just don't have the context of what our software business is doing and what are the vendors that…”
Hastings: Books inherited by Quanta from other providers contain widespread miscategorizations
“So when with Quanta inherit books from other providers, we see all of these things miscategorized, and it's actually very simple for us to get that understanding based on what we see as the memos and what we see as the receipt content and invoice content.”
Hastings: Bookkeepers spend most of their time manually moving CSVs
“I think the biggest example is just how much time is spent logging into a financial tool, downloading some sort of data, massaging it, and then uploading it into a traditional accounting system. That is so much of what bookkeepers are doing today.”
Hastings: Plaid is not accurate or sufficient for full bookkeeping
“So Plaid isn't a hundred percent accurate. I will say that we use a mix of Plaid and Finicity and other tools and the, but the best is when companies have APIs where the data is coming in? Well, the reason that Plaid isn't enough is that there's always small t…”
Hastings: QuickBooks Lacks Automated Checks and Requires Manual Data Verification
“And the way QuickBooks is set up is it just makes it really easy for you to make those mistakes, because it doesn't have all these checks built in. It's really just sort of a blank slate you can write into, and the only way to verify that the data is correct i…”
Hastings: QuickBooks Does Not Match Payroll Records with Bank Movements
“From my payroll system, I see that a payroll ran, and then in my bank, I see that the money moved to run that payroll. We match those things together and make sure that they're consistent, and QuickBooks does not do that, but we built that from the ground up i…”
Hastings: Waiting for monthly financial close is too slow for decision-making
“If you are waiting until the end of the month to understand what your finances are, the time has already passed to make the decisions that are important. You really need to be making decisions on a day-to-day basis, and that requires understanding the fundamen…”
Hastings: Hybrid software-services model is best for launching AI startups
“Well, that's one reason that the hybrid software and services model is a great way to get started on a business like ours because that last mile edge case, we do have humans do, and it's helpful for onboarding and understanding the, our customers' businesses, …”
Hastings: Founders must do manual dirty work to build great products
“And I think that, by the way, I think that's the best way to just build a great product. Just really see the dirty parts of it instead of just handing it off to someone else.”
Hastings: Quanta offers QuickBooks backup to ease customer migration fears
“Actually we do have a product that we call a QuickBooks backup, where if you, for whatever reason, want to do an export into QuickBooks or even keep a QuickBooks up to date, we support that. And it's a good way to get people in the door, especially we work wit…”
Hastings: Bench failure left customers without access to critical financial data
“You think of the companies like Bench going out of business and people have no access to their history of the most critical business data.”
Hastings: QuickBooks is still not fully cloud-based
“QuickBooks is still not even fully cloud, which is very funny.
I think, so a lot of people still use QuickBooks desktop.”
Hastings: Quanta only accepts clients that fit its automated software
“Our approach was we only work with companies that fit within the automation that we have built.”
Hastings: Pilot and competitors relied on large manual staffs using QuickBooks
“So pilots, any of the ones we just talked about, They have hired a very large staff that is just using QuickBooks. So they did not start by building their own QuickBooks replacement. They started using QuickBooks, and they started by hiring a lot of people, an…”
Hastings: Relying on manual labor prevents startups from transitioning to automation
“So once you hire the people to do it, and once you're supporting that, it's very hard to go the other way and come back to the automation.”
Hastings: QuickBooks' generic database approach prevents it from serving industries well
“QuickBooks, it can solve for any sort of business model because it's so generic, but because of that, it's sort of just, it's like I said, a blank slate, just a little database. You can write anything into there, and because of that, it can support any busines…”
Hastings: Bookkeepers misclassify LLM API costs into OpEx instead of COGS
“The past 15 years of SaaS businesses, everyone just assumes that their margins are incredible, but now companies are spending so much on the LLM API providers, and because traditional bookkeepers don't understand those new vendors and how to treat them, they o…”
Hastings: AI software margin profiles fundamentally differ from traditional SaaS
“Really becoming a requirement in the AI era because your costs are so volatile and changing every day, and the margin profiles of software businesses have completely changed with how expensive their costs are and how those costs scale when their users use them…”
Hastings outlines a 3-step playbook for AI automation startups
“I would say for anyone looking to use AI to start a company to displace what was traditionally done very manually by humans, first, see what they're doing really understand it then, two, do it yourself. As you bring on your first customers. So do the work your…”