Startup Spotlight: Acrely pivoted to home-services calls and found six customers in two weeks

Why it matters

Acrely's early story is a sharp example of customer pull shaping a startup's first market. Its six-customer claim is promising, but the later Lupitor brand and broader enterprise pitch make continuity, repeatable deployment and measurable customer outcomes the real tests.

Alek Stefanov and Derek Armfield said Acrely pivoted from real-estate scheduling to home-services calls and landed six customers within two weeks. The milestone, in Y Combinator's launch profile, captures Acrely's earliest public bet: build for businesses already asking for help.

The profile places Acrely in YC's Summer 2025 batch, and its launch post is roughly a year old. This is an account of an early customer-driven pivot, not a new September 2026 launch. The founders' public description offers a useful signal of customer pull, with a limit: six customers is a count, not evidence of recurring revenue, call volume, retention or the share of each customer's work handled by the software.

A change in customer, and in the job to be done

Acrely began with voice scheduling for real-estate agents, according to its YC launch post. Home-services businesses then came inbound, the founders wrote, and Acrely redirected its work toward HVAC, plumbing and electrical contractors. Its proposed agent would answer inbound calls, qualify customers, book jobs, dispatch technicians and follow up. Acrely said it integrated with ServiceTitan and other contractor CRMs, aiming to move a call through to a scheduled job instead of leaving the business with a recording or transcript.

That pivot changed more than the customer list. In the founders' account, they discovered a more defined operational task: connect a customer's phone call to scheduling and dispatch. The job depends on the agent understanding what the caller needs and then interacting with business systems. A fluent conversation alone would not complete it. This is an inference from the workflow Acrely described, not a claim that the launch profile independently tested those steps.

Acrely pitched that workflow against a familiar problem in home services: a call may arrive when office staff are busy, and a missed call can mean a missed booking. Its pitch was that an always-available agent could help contractors handle that first customer interaction without increasing office staffing. Acrely's launch materials did not quantify how many missed calls its customers had, how many calls its agents completed, or whether contractors reduced staffing. The six-customer figure describes adoption at one early point; it cannot establish those outcomes.

Two founders with different technical proof points

Stefanov and Armfield are described by YC as childhood friends from Pittsburgh who studied computer science at Carnegie Mellon. Stefanov had been the founding engineer at Caseflood.ai, a YC Winter 2025 company building legal AI agents. YC's profile also says he won a statewide Pennsylvania math olympiad and built Instagram accounts that reached millions of followers when he was 14. Armfield had worked on communications software for lunar rovers and reinforcement-learning controls for self-driving racecars, according to their YC launch account.

Those are distinct forms of experience, and they fit the product's two visible demands. The call agent needs to interpret a conversation, then reliably carry out an action in software used by a contractor. Stefanov's earlier work at Caseflood put him in a vertical-agent company; Armfield's described work involved communications and control systems. That connection is a reading of their backgrounds, not a claim the founders gave those experiences as the reason they chose home services.

The founders' shared Pittsburgh and Carnegie Mellon history also helps explain the speed of the pivot without making the two-week statistic do more work than it can. They already had a working direction in voice scheduling, according to the launch post. Their account says prospective home-services customers supplied the new use case. The post does not break out how many of the six customers paid, how deployments were configured, or whether all six came from that initial inbound demand.

The hard part is the workflow after the hello

For a contractor, answering a call is one step in a chain that can include identifying the service issue, collecting customer information, finding an available appointment, entering the booking and sending a technician. Acrely's early positioning put those downstream actions at the center. Its CRM integration claim matters because it implies the agent was meant to write into existing systems, rather than operate as a separate voice interface. The YC materials name ServiceTitan and other CRMs but do not document the depth of the integration or provide an independent test of booking and dispatch accuracy.

That distinction is also where the product faces a crowded category. Vapi markets enterprise voice-agent infrastructure for building and managing agents, while Retell AI describes a platform for call automation, integrations and task execution. Both companies' descriptions show that workflow actions and connections to customer systems are now part of the broader voice-agent pitch. Acrely's early case for standing apart was the contractor-specific sequence of call handling, booking and dispatch, rather than voice generation on its own.

Vertical focus can make the initial product easier to explain: a contractor wants a job booked and a technician dispatched. It can also make implementation demanding. Businesses may have different schedules, service areas, qualification rules and CRM setups. Acrely's launch claims do not say how much configuration each customer required or how much of that work could be reused. Those details would determine whether the product could scale as software or would remain dependent on custom setup for each operator.

Acrely's current site now carries the Lupitor name

The Acrely website now redirects to a site branded Lupitor, which describes a broader product for regulated enterprises. The Lupitor site presents agents across voice, chat, email, SMS and WhatsApp, and advertises cloud, sovereign-cloud and air-gapped deployment. Its listed integrations include Salesforce, Zendesk, SAP, Twilio and Oracle. That positioning spans beyond Acrely's original home-services voice agent into customer operations at banks, insurers, healthcare providers, telecommunications companies and public agencies.

The transition is visible in the domain and current site, while the public materials reviewed here do not establish when the branding changed or document the corporate relationship between Acrely and Lupitor. The two names should therefore not be treated as interchangeable evidence of a continuous product or customer base. The YC profile remains an account of Acrely's Summer 2025 company and early home-services pivot; Lupitor's current homepage describes what it markets today.

Lupitor's homepage also displays a claim that it raised $5 million led by TQ Ventures. That is a financing claim made under the Lupitor name. It should not be added to Acrely's funding history on the strength of the redirect alone. The same homepage says Lupitor is in production at Fortune 500 enterprises and trusted by more than 40 enterprises, and it advertises metrics such as a 72% containment rate for a European banking group. Those are company-published claims; the page does not provide enough detail in the displayed figures to assess their measurement periods, denominators or independent verification.

What the early traction does, and does not, show

A two-week pivot followed by six customers is a credible early sign that Acrely found a use case customers wanted to explore. It does not establish product-market fit by itself. The count leaves important business questions unanswered: whether customers kept the product, paid enough to support the service, or used it for a meaningful portion of incoming calls. For workflow software, the count of logos matters less than the repeatability of deployment and whether the agent reliably completes the task the customer cares about.

Still, the founders' early decision offers a clear operating lesson. They began with a defined voice product, listened when a different customer group came forward and shifted toward a workflow with a concrete business result. Acrely's launch material made that sequence unusually legible: real-estate scheduling became contractor call handling, and contractor call handling was tied to booked work and dispatch. The evidence supports a focused early customer-discovery story. It does not support treating six early customers as proof of scale, or Lupitor's newer enterprise claims as a verified continuation of those six accounts.

Acrely's most durable early insight may be that the phone call is only the entry point. Businesses buy the completed task: an appointment on the schedule, the right technician assigned, and a customer who knows what happens next. The shift from Acrely's original home-services framing to Lupitor's broader enterprise positioning suggests the founders are now selling a wider version of the same operational premise. The available public materials show the change in positioning; they do not establish how much of the original contractor product survives inside it.