Self-hosted AI phone agents for the enterprise — $106M raised after 180 investor rejections
HQ: San Francisco, CA, United States | Founded: 2023 | Employees: ~50-100 | Stage: Series C ($50M led by Dell Technologies Capital, June 2026); $106M+ total raised | Website: https://www.bland.ai
Bland AI is an enterprise platform for AI phone agents: software that makes and receives calls with natural conversation, completes tasks in business systems, and — critically for its customer base — runs on infrastructure Bland owns and can deploy inside the customer's environment. Its API and tooling let enterprises build agents for support, sales, collections and operations, with call analytics and human handoff built in.
The company was founded in 2023 by Isaiah Granet and Sobhan Nejad, who had previously built and sold a dating app in an eight-figure exit. Granet, a Washington University in St. Louis alum, came to voice AI convinced that the phone — not chat — was where enterprise automation would actually pay. Early fundraising was brutal: 180 rejections before the pre-seed, answered with a $60K billboard on Highway 101 advertising Bland's phone agents — a stunt that became the company's calling card. The momentum that followed was real: a $16M Series A led by Scale Venture Partners in August 2024, a $40M Series B led by Emergence Capital in January 2025 (pre-seed to Series B in roughly a year), and a $50M Series C led by Dell Technologies Capital in June 2026, bringing total funding past $106M.
Bland's technical positioning is distinctive: rather than orchestrate rented models, it builds and runs its own voice infrastructure — its own hardware and self-trained voice components — enabling self-hosted deployments for customers with strict data requirements. Enterprises use Bland for high-volume customer support, appointment scheduling, payment reminders and outbound campaigns, running millions of calls with analytics on every conversation.
Bland AI is the contrarian infrastructure bet in voice AI: while rivals rent model APIs and cloud GPUs, Bland owns its stack — self-built voice models and its own hardware, with self-hosting options for enterprise customers who won't let conversations touch shared infrastructure. That's expensive and technically brutal, and it's the pitch: banks, healthcare systems and telecoms that won't buy voice AI any other way. A $16M Series A (Scale Venture Partners, Aug 2024), $40M Series B (Emergence, Jan 2025) and $50M Series C (Dell Technologies Capital, June 2026) — $106M+ total — priced that thesis.
The founder story is Silicon Valley folklore in the making. Isaiah Granet and Sobhan Nejad had already built and sold a dating app for eight figures when they started Bland; Granet then collected 180 investor rejections before the round closed, responded by buying a $60K billboard on the 101 to advertise phone agents to the VCs who'd passed, and rode the stunt into one of the fastest pre-seed-to-Series-B climbs of the 2024-2025 cycle. The follow-through mattered more than the stunt: production deployments running millions of calls, an enterprise security posture competitors can't match, and hardware economics that improve as volume grows.
The risks are the voice-infrastructure standard set — model labs improving native speech, better-funded rivals, and the regulatory glare on AI phone calls — plus the open question of whether owning hardware is a moat or a millstone as cloud voice costs fall. But if enterprises end up requiring self-hosted, auditable voice AI, Bland built for that future first.
Browse: Home | Blog | About | All Startups A-Z | View full profile on StartupWiki