YC Root AccessThis Is The Next Industry AI Will Disrupt
CHAPTERS
- 0:00 – 0:58
Why Big Four dominance is vulnerable to AI-driven disintermediation
Dominic argues that accounting firms have become entrenched middlemen and that AI can change how accounting outcomes are delivered. He frames the shift as a “tectonic” change that could make the traditional firm model far less relevant.
- •Accounting firms as entrenched intermediaries rather than essential value creators
- •AI as the catalyst for a structural shift in delivering accounting/tax outcomes
- •Skepticism that Big Four have “earned” long-term reverence
- •Vision of a post–Big Four world becoming plausible
- 0:58 – 1:05
What accountants actually do day-to-day: the spreadsheet labor reality
Tom asks for a computer-science-level description of accounting work, and Dominic describes early-career work as largely manual spreadsheet manipulation. He illustrates how much of the job is copying, transforming, and calculating rather than true expert reasoning.
- •Junior accounting work as spreadsheet-based data movement and arithmetic
- •Manual workflows disguised as professional expertise
- •The work is often repetitive and structured enough to automate
- •The gap between perceived complexity and actual execution
- 1:05 – 2:37
R&D tax credit case study: interviews, rubber-stamped estimates, and manual write-up
Dominic recounts how R&D tax credits were built from hours of employee interviews that produced questionable time estimates. The process then becomes a manual translation of notes into spreadsheets to compute the credit.
- •Long on-site interviews to elicit time allocation estimates
- •Incentives to inflate eligible time percentages
- •Manual note-taking converted into spreadsheet calculations
- •Illustration of process inefficiency and fragility
- 2:37 – 4:18
The real bottleneck: substantiation and proof, not math
The conversation shifts to why accountants exist: not to do arithmetic, but to prove what happened. Dominic explains that documentation (Jira tickets, Git issues, logs) is central, and AI’s ability to read/organize evidence changes the game.
- •Accounting/tax work depends on defensible evidence and documentation
- •‘Contemporaneous documentation’ is the compliance cornerstone
- •Software companies have richer proof trails than many offline industries
- •AI unlocks automation by understanding unstructured text and artifacts
- 4:18 – 6:26
From RPA theater to frontier models: why automation works now
Dominic contrasts earlier automation attempts (RPA tools, certifications, licenses) with today’s frontier LLMs. He claims modern models now match or exceed junior and mid-level staff, and are approaching senior-expert capability for many tasks.
- •Past ‘automation’ efforts were superficial and low-impact
- •ChatGPT-era models changed feasibility dramatically
- •Dominic’s early exposure to GPT-3 hinted at the coming inflection
- •Models improve continuously—“today is the worst they’ll ever be”
- 6:26 – 7:47
Founder backstory and Onshore’s mission: automating tax, then audit and advisory
Dominic explains his accounting + computer science background and his experience at Grant Thornton. He defines Onshore’s aim: automate accounting-firm work products using AI, reducing repetitive labor while improving outcomes and traceability.
- •Intersectional background: accounting + CS as a strategic advantage
- •Grant Thornton experience revealed execution-heavy, low-leverage work
- •Onshore automates tax work now, targets audit/advisory later
- •Focus on outcome quality, not hours billed
- 7:47 – 11:10
Why incumbents resist: billable hours, partner incentives, and ‘AI spend’ as PR
Dominic argues big firms publicly tout AI investments but rarely transform core workflows. He attributes resistance to the collapse of hourly billing as a value metric and to senior partners’ short time horizons and pension incentives.
- •Incumbent AI initiatives often amount to license purchases with little adoption
- •Lack of in-house engineers blocks real product development
- •Automation threatens the billable-hour value narrative
- •Mandatory retirement and payout structures discourage long-term reinvestment
- 11:10 – 13:37
The failed go-to-market: selling automation to accountants who fear fewer hours
Dominic details Onshore’s initial approach: sell productivity software to accounting firms. The pitch appealed to partners (higher margins) but threatened staff (fewer billable hours), causing internal resistance that stalled adoption.
- •Early traction selling to firms, but persistent implementation friction
- •Partners liked margin expansion; staff feared job/billable-hour loss
- •Cultural preference for ‘our way’ and edge-case objections
- •Conclusion: accountants are among the least motivated buyers of automation
- 13:37 – 15:55
Why legal AI adoption differs: perceived risk and shifting billing models
David compares accounting to legal, noting legal AI companies are succeeding. Dominic points to perceived stakes (legal errors feel existential) and suggests lawyers may be more open to change, while billing structures are evolving toward project fees.
- •Legal work carries higher perceived downside, sustaining premium pricing
- •Law firms have been moving toward project-based billing
- •Accounting often still anchors fees to implicit hourly economics
- •Cultural openness to change differs across professions
- 15:55 – 17:45
The pivot: firing firm customers and selling directly to corporations (end users)
After years of resistance, Dominic reframes the value chain: corporations benefit, not accounting firms. Onshore terminates firm contracts, rebuilds for taxpayers/CFOs, and launches an outbound hustle to sell the outcome directly.
- •Re-identifying the true customer: the company/taxpayer
- •Accountants positioned as an ‘artificial middleman’ in many workflows
- •Bold reset: terminate existing customers and rebuild product focus
- •Direct sales via cold outreach to corporate stakeholders
- 17:45 – 18:46
Early proof and YC entry: the Miami deal and rapid acceleration
Dominic describes landing an early corporate customer through cold DMs and an in-person meeting, even while the team was cash-constrained. During the same trip they receive and complete a YC interview and get accepted, accelerating the company’s trajectory.
- •First corporate win driven by trust, efficiency, and lower price
- •Old-school closing: paper contract signed in person
- •YC acceptance as a pivotal moment for the company
- •Validation that direct-to-corporate sales motion can work
- 18:46 – 21:44
Co-founder credibility and loss: building with a senior partner from Grant Thornton
Dominic explains why he recruited Mark, a senior partner, to add credibility and domain expertise. He reflects on Mark’s instrumental role in early customer trust and product realism, and shares the impact of Mark’s passing in 2024.
- •Credibility advantage in a ‘no one gets fired for buying IBM’ market
- •Domain mastery: knowing what was delivered and how it was sold
- •Unusual founder pairing (junior associate + senior partner) that worked
- •Mark’s death during a pivotal fundraising period
- 21:44 – 25:34
Forecasting the new accounting pyramid: fewer juniors, more leverage, different roles
The group explores how accounting changes if Onshore and similar companies succeed. Dominic expects a major reshaping: less junior-heavy pyramids, more technical professionals, and AI agents absorbing routine work while humans shift to higher leverage tasks.
- •A post–Big Four world becomes thinkable as outcome delivery changes
- •Revenue per employee rises sharply as AI removes low-level labor
- •Firm structure shifts: fewer juniors, more sales/expertise/engineering at top
- •Work may move partially in-house at corporations using better tools
- 25:34 – 28:41
Onshore’s scale economics and wedge strategy: start narrow, expand via trust
Dominic shares growth targets (toward $100M revenue) with a small headcount and contrasts this with incumbent efficiency. He argues the wedge approach—dominate R&D credits first, then expand based on customer pull—is superior to trying to ‘boil the ocean.’
- •Targets: rapid revenue growth with ~60–100 employees
- •Order-of-magnitude better revenue per employee than legacy firms
- •Incumbents worsen efficiency by offshoring rather than retooling
- •Wedge strategy: win one workflow, then expand via customer-driven roadmap
- 28:41 – 31:29
Beyond Onshore: the next disruption target might be Excel for knowledge work
Asked what he’d build next, Dominic suggests reimagining Excel’s role in professional services. He argues spreadsheets are overloaded with workflows they weren’t designed for, and there’s room between Excel and ‘vibe coding’ tools for accessible, higher-level systems.
- •Excel as foundational but overextended infrastructure for many industries
- •Gap between tools like Excel and developer-centric platforms like Replit
- •Barriers: awareness, incentives, and agency for non-technical workers
- •Opportunity to consolidate macro-workflows into fewer, smarter tools
- 31:29 – 33:55
Adoption curve: capability is here; acceptance is catching up; jobs shift later
They close by comparing AI’s rapid capability gains with slower labor-market changes. Both expect broader acceptance that AI can do knowledge work soon, while employment impacts may take several years to materialize.
- •Models meet/exceed expectations; workplace change lags behind
- •Professional acceptance of AI is increasing rapidly
- •Customers increasingly want to ‘see the AI’ in action
- •Expect delayed employment effects over a 5–7 year horizon
