🌏 中文版
Funding Details
| Item | Value |
|---|---|
| Company | Broccoli AI (San Francisco, US) |
| Round | Series B |
| Amount | $30M |
| Lead | Khosla Ventures |
| Participants | Y Combinator, Operator Partners |
| Valuation | ~$300M (per a person familiar with the matter; the company has not confirmed; prior-round valuation undisclosed) |
| Total raised | $53.5M ($1.5M YC seed + $22M Series A + $30M Series B; the Series A was disclosed publicly for the first time alongside this round) |
| Founded | 2024 (founders arrived in San Francisco in January 2024, after an earlier electric-vehicle API startup) |
| Headcount | 150+ (up from about 15 a year ago, a 10x increase) |
What the company does
Broccoli AI builds front-desk automation agents for home-service businesses — plumbing, HVAC, and electrical contractors. It isn't selling a chatbot; its agents answer inbound calls, book jobs directly into the scheduling software a contractor already uses, follow up on open estimates, and remind customers to renew service memberships.
Its differentiator is not asking owners to change their habits: Broccoli plugs its agents into existing scheduling and bookkeeping systems rather than requiring a switch, handling the chores owners complain about most — nobody picking up the phone after 5 p.m., or no one following up on estimates technicians wrote. More than 1,500 businesses now use it, including Ace Hardware Home Services, Mr. Electric, and Any Hour Group, and in the past seven months it has expanded from plumbing/HVAC/electrical into roofing, pool care, garage doors, and home improvement.
Co-founders AJ Jain and Aditya Ketkar met as students at BITS Pilani in India. Before arriving in San Francisco in early 2024, they had tried several ventures that never scaled — a campus delivery app, an interview-prep bootcamp, a universal EV API in Europe. In May and June 2024, the pair walked into roughly 140 contractors' shops across Northern California carrying notebooks instead of laptops, eventually landing their first customer at Armstrong Plumbing in Sacramento — owner Marcelle Flowers later became Broccoli's COO. The company says its AI now handles about 80% of tasks, equivalent to roughly 30% of a typical contractor office's staff workload, with the remaining 20% — emergencies like a flood or a fire — still routed to humans.
What this round signals
What it means for the agent ecosystem
Broccoli shows a vertical agent doesn't need new infrastructure to win — it plugs directly into the scheduling and bookkeeping software owners already run, acting as a "ghost office employee" rather than replacing the whole stack. More notable is the founders' approach: instead of writing code first and finding customers later, they spent months embedded inside a plumbing shop, trading direct observation for domain knowledge before building anything. That suggests the next edge for vertical agents may not be the model itself, but whether founders are willing to spend the time to actually understand how a traditional industry runs.
What investors are betting on
Khosla Ventures personally led both rounds — Series A and Series B — just nine months apart, signaling its conviction in the growth curve never wavered even after the first round. Khosla has long favored markets where "the long tail of small businesses is large but software penetration is low" — home services in the US is a market worth hundreds of billions of dollars, and most owners haven't even adopted cloud scheduling software yet, leaving far more room for an AI agent to break in than in enterprise software markets SaaS has already saturated.
Numbers worth watching
Figures compiled from Fortune's reporting (2026-10-28); axis starts at 0, domain 0-32 (US$M), k=(300-60)/32=7.5, bar top y=300-amount x 7.5.
- Headcount grew from 15 to over 150 in a year — a 10x increase that outpaces the typical growth rate at Series B stage
- Revenue has tripled since the Series A (absolute figures undisclosed); Y Combinator separately disclosed that Broccoli's annualized booking run rate already exceeds $1 billion, with ARR projected to top $60M by year-end
- $53.5M in total funding bought roughly a $300M valuation — a multiple of about 5.6x, notably conservative compared to the 20-50x multiples common in this AI funding wave, reflecting a more cautious valuation logic for labor-intensive vertical agents with margin structures closer to traditional SaaS
Watchlist status
Broccoli AI isn't on the watchlist yet. None of the 296 companies currently tracked maps cleanly to "AI agents for home-service businesses" — recommend adding it under the closest fit, section D11 (Workflow Automation), tracking: front-desk agents for home-service contractors (plumbing/HVAC/electrical) handling calls, scheduling, and estimate follow-up, with Khosla Ventures leading both Series A and B.
Today's takeaway
Both Broccoli founders had tried and failed at several ventures before this one, each time betting on "smarter technology" to win. It wasn't until they set their laptops aside and spent three months embedded in a plumbing shop that they found something that could actually scale. That's a reminder that a vertical agent's real moat usually isn't model capability — anyone can call the same APIs — it's whether anyone is willing to spend the time to truly understand the daily grind of a traditional industry. That can't be bought with more compute; it only comes from time spent.