CHAPTERS
- 0:05 – 0:59
Warp’s traction and what the conversation will cover
Harj introduces Ayush Sharma and Warp, framing the company as an AI-native employee management platform with significant payroll volume and rapid growth. He tees up the discussion around Warp’s origin story and what “AI-native” really means in practice.
- •Warp positioned as AI-native employee management for high-growth companies
- •Scale: 1,000+ customers; $600M payroll processed last year; aiming for $2B next 12 months
- •YC W23 and newly announced $60M Series B
- •Episode focus: origin story + meaning of AI-native product/company
- 0:59 – 2:52
From a small town in India to MIT via a physics obsession
Ayush recounts growing up in a small town in India, initially aspiring to be a theoretical physicist. He describes the pivot from the expected IIT track to applying to MIT, where he studied CS/math/physics and focused on machine learning.
- •Early passion for physics; self-driven learning with Feynman lectures
- •Family and societal expectations pushing toward engineering/IITs
- •Behind-the-scenes pivot to apply to MIT (and Columbia)
- •MIT experience and ML specialization; notable “first from my state” milestone
- 2:52 – 4:05
Early startup experimentation: consumer “tarpit” ideas and the pivot point
Harj and Ayush discuss Ayush’s first attempts at consumer/social apps and why they ultimately didn’t feel like venture-scale opportunities. The key inflection was shifting from building something fun to systematically identifying painful problems they personally experienced.
- •Typical early founder path: social/consumer app experimentation
- •Initial traction/community, but lacked a path to a massive outcome
- •Team exercise: list real problems they’d faced and would solve for themselves
- •Set the stage for moving from consumer to B2B infrastructure
- 4:05 – 5:39
Choosing an “unsexy” problem: payroll compliance pain as the spark
Ayush explains the personal frustration of setting up payroll-related compliance (e.g., state withholding and labor accounts) while juggling startup fires. The timing (2022–2023) and his ML background made it compelling to attempt automation with LLMs/agents.
- •First-hand pain: setting up multi-state compliance accounts was brutally time-consuming
- •Belief that the process should be automatable end-to-end
- •LLMs/agents emerging as a new toolset to tackle messy workflows
- •Warp begins as an attempt to apply AI to payroll/compliance complexity
- 5:39 – 6:59
Schlep blindness and the bet that messiness hides big opportunities
Harj presses on motivation: why commit to a seemingly boring domain. Ayush argues that “unsexy” and complex problems are often under-attacked, citing schlep blindness and examples like Stripe’s early banking/API grind, and connects payroll to a bigger employee-management platform vision.
- •Motivation partly came from the problem being “unsexy” (less competition)
- •Schlep blindness as a founder advantage in messy domains
- •Validation: many others share the same pain
- •Vision expands: payroll compliance as the foundation for a broader platform
- 6:59 – 10:03
The wedge that started Warp: multi-state payroll, post-COVID complexity, and ‘one employee triggers compliance’
Ayush recounts the YC interview pressure-testing whether the multi-state wedge is deep enough in a competitive market. He explains why the trend is favorable: distributed work increased multi-state needs, and payroll tax compliance triggers with just one employee—making complexity explode early for scaling companies.
- •Competitive landscape acknowledged; wedge questioned heavily during YC interview
- •Trend shift: multi-state employment became common post-COVID
- •Key insight vs sales tax: payroll compliance starts with a single hire in a jurisdiction
- •AI/agents enable automation that traditional software + headcount couldn’t match
- 10:03 – 12:31
What ‘AI-native’ means for Warp’s product direction: serving the whole company lifecycle
Warp’s scope expands because customers—especially startups without large HR/legal/accounting teams—push for more capabilities as they scale. Ayush explains that Warp aims to be the platform that prevents ops teams from scaling linearly with company growth.
- •Early adopters (startups) shaped the roadmap as they grew from 5→200+ employees
- •Need to go beyond payroll into broader employee management to retain customers across stages
- •AI-native goal: make core ops/compliance work agent-executable
- •Promise: help high-growth companies avoid linear hiring in HR/finance/ops
- 12:31 – 14:36
AI-native vs traditional HR tech: automation changes company structure, not just features
Ayush contrasts incumbents—often staffing 30–40% of headcount in support/compliance/ops—with Warp’s approach. He highlights Warp’s ability to cover all U.S. jurisdictions with a tiny tax team, arguing AI enables a fundamentally different operating model.
- •Traditional compliance-heavy SaaS scales via headcount (support, tax, ops, legal)
- •Incumbents often allocate 30–40% headcount to these functions
- •Warp’s claimed efficiency: 1,000+ customers and nationwide coverage with ~1–2 tax staff
- •AI-native framing extends to org design and operating leverage
- 14:36 – 16:42
Why AI favors technical founders—and why retrofitting incumbents is hard
Ayush argues that pre-AI SaaS favored distribution and sales-led advantages, while AI shifts leverage toward technical founders who can build new primitives. He also claims incumbents struggle to bolt AI onto legacy architectures, and points to market anxiety around older enterprise platforms.
- •Pre-AI era: advantages skewed toward distribution-heavy, sales-oriented companies
- •AI shift: technical founders can build net-new primitives without legacy assumptions
- •Retrofitting AI as “thin chatbots” on legacy systems is insufficient
- •Incumbent risk illustrated via enterprise market sentiment (e.g., Workday)
- 16:42 – 19:29
Systems of record: shared truth, defensibility, and the race to become the new default
Harj brings up ‘systems of record’ as potentially defensible in an agent-driven future. Ayush defines them as shared-truth databases tied to business processes (e.g., Salesforce), notes switching/integration gravity, and frames the competitive race as AI-native systems vs incumbents layering AI on top.
- •System of record = shared truth + logs/history + process semantics
- •Example: Salesforce’s defensibility strengthened by integrations and workflow gravity
- •Not all systems of record are equally defensible
- •Strategic race: AI-native upstarts becoming the next system of record vs incumbents adapting
- 19:29 – 21:29
From systems of record to systems of intelligence: agents swarming enterprise software
Ayush warns incumbents may degrade into “dumb data stores” while value moves to the agent-orchestration layer. He argues the next generation combines record-keeping with native agent execution, guardrails, permissioning, and trustworthy automation—anticipating an enterprise ‘flip’ where agents dominate usage.
- •Incumbent risk: becoming database-only while orchestration layer captures value
- •Next-gen vision: system of record + built-in agentic intelligence and controls
- •Guardrails/permissioning critical for sensitive HR/payroll actions
- •Agent traffic trend suggests enterprises will soon see similar ‘flippening’
- 21:29 – 23:28
Why investors moved fast on the $60M Series B: the AI-native HCM narrative crystallizes
Harj asks what drove Battery to invest with high conviction; Ayush says the round was a preempt less than 10 months after Series A. He attributes speed to a shifting market view: many enterprise categories have AI-native challengers, but HCM lacked a clear winner—Warp increasingly fits that slot.
- •Series B as a preempt shortly after Series A
- •Investor pattern-recognition: AI-native challengers emerging across enterprise categories
- •Claim: AI-native HCM remains one of the last major categories without a clear leader
- •Momentum: product shipping + customer growth + narrative shift combined
- 23:28 – 25:15
What Warp will build next: full employee platform + trusted agent infrastructure
Ayush explains the capital will fund two parallel priorities: expanding a broad, engineered employee management platform and building agent infrastructure that can safely act on sensitive employee/payroll/tax data. He lists major product lines already underway, from benefits brokerage to device management and global contractors.
- •Two simultaneous bets: platform breadth + agentic infrastructure
- •Platform scope: benefits brokerage (licensed nationwide), device management, contractor tools, HRIS basics, integrations
- •Need for reliability and trust when agents act on payroll/tax/employee data
- •Funding used to scale engineering and product for the next phase
- 25:15 – 26:58
Warp’s upcoming customer agent: natural-language workflows for onboarding and compliance
Ayush previews the GA launch of Warp’s customer-facing agent, extending AI from back-office automation to direct user control. The promise is that anything doable via clicks can be done via natural language—enabling complex, multi-step onboarding and compliance workflows without writing code.
- •Shift from background agents to customer-facing agent capabilities
- •Parity claim: “anything you can do by clicking buttons” can be done with the agent
- •Examples: jurisdiction-specific onboarding + benefits + leave policies + device/app provisioning + training
- •Natural-language workflow creation replaces complex rule builders; early customer use cases emerging
