Podcast Summary: Startup Stories - Mixergy
Episode #2288: She had people do AI’s work
Host: Andrew Warner
Guest: Helen Hastings, Founder of Quanta Accounting Software
Release Date: December 17, 2025
Episode Overview
This episode dives into the innovative journey of Helen Hastings, founder of Quanta Accounting Software, who pioneered a hybrid approach using both human expertise and AI to streamline bookkeeping and accounting for startups. Andrew Warner explores Helen's unique methodology—starting with manual processes, deeply understanding user pain points, and “copying the bookkeeper’s brain”—enabling Quanta to raise $20 million and win a rapidly growing segment of early-stage and growth SaaS companies.
Key Discussion Points & Insights
1. Starting by Having Humans Do AI's Work
- Helen began Quanta by having humans perform tasks that were eventually automated by AI (00:00–01:35).
- Example: Early on, people manually read memos and invoices to correctly categorize expenses (e.g., Amazon purchase vs. AWS bill), a task now handled by LLMs.
- Quote:
"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."
— Helen (00:37)
2. Deep User Research: Shadowing, Networking, & Hustle
- Helen intensively shadowed bookkeepers across various industries, leveraging her network and cold outreach via LinkedIn (02:25–05:48).
- Silicon Valley's openness enabled her to sit in with competitors for learning—not common elsewhere.
- Her background (Stanford, Google, NerdWallet, Affirm) added credibility but hard work and persistence were central.
- Quote:
“Once I got in the groove of it, I really don't know how I would have started this company if I hadn't done that.”
— Helen (04:14)
3. Discovering Opportunities for Automation
- Observed common pain points:
- Manual downloading/uploading of financial data into accounting systems, even in “automated” environments using tools like Plaid (05:48–07:06).
- Many popular tools (QuickBooks) were a core source of friction due to their generic nature and lack of robust automation.
- Redundancy and continuous reconciliation are built into Quanta, ensuring accuracy (09:44–11:14).
4. Building vs. Integrating with Existing Platforms
- Quanta opted not to layer on top of QuickBooks/Xero, but to rebuild the accounting stack from the ground up, overcoming legacy limitations (07:38–09:44).
- Quote:
"Building financial systems of record is the word that we use. That was my specialty and that was what I saw as the opportunity."
— Helen (07:38)
- Quote:
5. Maintaining Data Integrity
- Existing products let users make unchecked edits, breaking the source of truth (08:48–09:56).
- Quanta enforces correct edits and traces every change, maintaining reliable, clean data essential for automation and AI use.
6. The Customer Problem: Speed & Visibility
- Traditional bookkeeping is slow—often by weeks/months, making data obsolete for startups (13:40–15:59).
- Quote:
"Once you have it available to you and close to real time, you're actually looking at it and using it to make decisions."
— Helen (14:28) - Quanta transitions bookkeeping from a compliance/tax “chore” to a strategic, real-time dashboard.
7. Customer Discovery & Early Validation
- Helen strongly advocates talking to real customers and testing willingness to pay before building the product (16:25–17:13).
8. The Hybrid Approach: Bridging the Manual “Last Mile” with Humans
- To handle edge cases, Quanta used humans (including engineers and Helen herself) in the loop, while automating 90%+ of standard workflows (17:30–19:43).
- This hybrid model built confidence with users, especially during onboarding.
9. Overcoming Trust Barriers
- Accounting buyers are skeptical of new tech—Quanta offers a “QuickBooks backup” so hesitant users can keep their prior system current as a safety net (20:11–21:59).
- Quote:
"The trust problem... It's so enormous in accounting and, and we still get it every single day."
— Helen (20:11)
10. Why Incumbents Endure—and Quanta’s Wedge
- QuickBooks dominates (especially in the U.S.) due to its generic, all-purpose approach; Xero strong internationally (22:47–24:35).
- Helen attributes Quanta’s potential for success to strict focus: only serving companies whose financial workflows can be completely automated—mainly early/growth stage software & services without physical inventory (25:55–31:21).
- Others failed by going too horizontal, hiring armies of bookkeepers and depending on QuickBooks.
11. The AI Moment: Why Now?
- The AI wave opens minds—companies realize their old software is outdated. Quanta leverages this openness and also helps startups track their now-volatile, usage-based margins (32:17–35:06).
12. The Importance of Owning & Structuring Data
- Clean, structured data is essential to reliable AI automation and analytics (35:06–35:51).
- Quote:
"If you're taking ChatGPT and putting on messy data, you can only get garbage out if you put garbage in."
— Helen (35:51)
- Quote:
Notable Quotes & Memorable Moments
-
On Silicon Valley’s openness:
"I mean, I chat with competitors and people believe in just the more we share, the, the more everyone will learn and get better."
— Helen (03:09) -
On saying 'no' as a strategy:
"Our approach was we only work with companies that fit within the automation that we have built... And we still do. Even just earlier this morning... that discipline has made sure that we actually have a viable business and we are operating at a quality level that you can't find elsewhere."
— Helen (25:55) -
On legacy products:
"The only way to verify that the data is correct is people manually looking at it. Versus, the thing that we've built from the ground up is checking itself all the time."
— Helen (09:56) -
On trust in bookkeeping software:
“It’s almost too good to be true. So it's, it's, this is a big problem in, in the AI era I think is these too good to be true problems.”
— Helen (20:11) -
On starting companies for the AI age:
"First, see what they're doing, really understand it. Then two, do it yourself as you bring on your first customers... And then three, make sure you are in your data storage, really understanding and building up a clean, structured data. Because when you apply AI on top of that, it can be like magic."
— Helen (35:51)
Timestamps for Key Segments
- 00:00–01:35: Helen describes starting with humans doing AI's work; why language understanding in accounting matters.
- 02:25–05:48: Shadowing bookkeepers and the importance of hands-on research.
- 05:48–07:06: Discovery of major automation opportunities in traditional accounting workflows.
- 07:38–09:44: Why Quanta was built from scratch instead of on QuickBooks/Xero.
- 13:40–15:59: Traditional bookkeeping’s time lag and impact on decision making.
- 16:25–17:13: The importance of validating customer willingness to pay.
- 17:30–19:43: Quanta’s hybrid model for handling edge cases and building trust.
- 20:11–21:59: Dealing with the trust problem—offering a QuickBooks backup.
- 25:55–31:21: Why Quanta is disciplined about its user focus and how it learns from failed competitors.
- 32:17–35:06: The impact of AI on buyer mentality and the changing needs of SaaS businesses.
- 35:51: Helen’s advice for building AI companies: start manual, structure data, automate.
- 36:46–37:05: Fundraising update: $20M raised; ~100 customers.
- 37:13–39:56: Announcing Prism: AI answering any finance question, with full audit trail.
- 40:04–41:02: Helen on what makes Elad Gil and his team (investors) special.
What's Next for Quanta?
- The newly launched Prism product enables founders and operators to ask any question about their business in natural language, receive clear answers, and view the source trail for verification (37:13–38:41).
- Example Questions: “Which customers churned last month?” “Which products are most profitable?”
- Continuing expansion focused on early-stage/growth SaaS & service businesses, adding integrations, with eventual plans to address new verticals like inventory.
Actionable Lessons for Startup Founders
- Do deep user research and shadow real customers.
- Start with humans in the loop: manual + software > pure software until automation is robust.
- Be disciplined about your initial segment; only automate what you can truly support.
- Maintain clean, structured data—critical for good AI.
- Adopt a hybrid model and address buyer trust by providing continuity and transparency.
Final Thoughts
Helen’s story is a playbook for building a high-trust, AI-first product in a risk-averse, heavily regulated market. Her disciplined focus, hands-on approach, and willingness to “do what doesn’t scale” set Quanta apart—a compelling example for founders aiming to automate traditionally human workflows.
[For further details, visit Quanta at usequanta.com and check out Quanta's integrations page for supported tools. Helen raises the bar for founder research and product iteration in the AI era.]
