Phone calls suck. So, we bought a giant telephone car, raised $65,000,000, and now we're fixing them with AI. 🧵

Salesforce just acquired Fin AI & with it 15 years of customer data. We prefer to keep your data in the same place you asked for it, always.

Our model’s still up, incase anyone is wondering.

Hearing is pretty important (for your voice agent). So we fixed it. Introducing: Noise Cancellation v2 #bland

Most noise cancellation is a filter. Hunt for bad frequencies, strip them out. Bland Noise Cancellation is smarter. It finds the one voice that matters, isolates it, and rebuilds it before the transcriber ever hears a thing. Your callers aren't in silent bubbles. They're on the road, in lobbies, in the real world. So we built AI for that. Live in Bland. Try it now → app.bland.ai.

The golden rule of AI is garbage in, garbage out. But what if less garbage in? With AI voice agents, environmental noise (road noise, lobbies, coffee shops, etc) triggers false interruptions. When it’s loud enough, agents go silent. We just shipped V2 of Bland’s noise cancellation system. Tested across 3,744 live production calls: 16% more accurate transcription with noise cancellation enabled. It filters noise before it ever reaches the model. Less garbage in! Hear for yourself: app.bland.ai

The most important conversations in fintech aren't happening at conferences: they're going on behind the scenes in small rooms with the right people. Join us June 11 to hear from folks directly disrupting the industry: Victor Cardenas Codriansky (CEO, Slash) and Isaiah N. Granet (CEO, Bland), along with a special guest for a candid panel on AI in financial services, will talk what's changing, what the challenges are, and what it takes to actually deploy in a regulated industry. Stay after the panel for a networking hour with drinks and bites! Spots are limited, register here: luma.com/9dyfyaox

A lot of people claim to have AI live. We are proud to prove it. Our impact for one customer: $430,000,0000 Four hundred thirty million Read: bland.ai/case-studies/americ…

Norm for iMessage is here. 🔵 Norm lets you get anything done in Bland via text. It can fix your pathways, run tests, pull reports, and send daily reminders. Want to see how it works? Text 1-321-424-0172 and tell Norm to build you a phone agent. Book a demo if you want to see how you can build your own iMessage agents in Bland. bland.com/for-enterprises

iMessage agents are here! Turn your AI agent into a branded contact in every customer’s phone, not another anonymous SMS number. With Bland iMessage, agents can: - Send & receive photos, videos, PDFs, spreadsheets, and voice memos - Use tapbacks, stickers, and native iMessage features - Show typing indicators, read receipts, and blue bubbles - Communicate with end-to-end encryption We built an iMessage version of Norm. Everything you normally do in Bland, can now be done via text. Text 1 (321) 424-0172 to: - Build or troubleshoot agents - Send reports or daily reminders - Query analytics Want to build your own iMessage agent? Book a demo here: bland.com/for-enterprises

One of our engineers has Zapier making calls for him. Zapier SDK Hermes Agent Bland Now he just messages his bot in Telegram: "call the pharmacy, see if my script is ready" and it makes the call (pending human approval, of course) Just point your agent at the @Zapier SDK. Connect to 9,000 apps with no extra setup. Next up: Zoom meetings? 👀

excited to announce our long held strategy of moving into an athleisure brand

Multiplayer editing is now live! Multiple teams of people can work on the same pathway at once.

We benchmarked our new transcription model (Fluent) across 250 hours of real-world audio. 5.9% WER — vs 8.1% from the leading real-time voice AI provider and 6.5% from OpenAI's Whisper. That's a 27% reduction in transcription errors. For voice agents, fewer errors = fewer misunderstood requests = more calls that resolve on the first pass. Learn more: bland.ai/blogs/fluent-next-g… Give it a try at app.bland.ai

Agent < > Agent testing is live - Build from prompts - Build from past calls - CI / CD deployments - Norm can automatically and recursively fix broken unit tests

npm i bland-cli bland-dev use claude opus 4.6 to create ai agents, right from your terminal.

we built a dev terminal for bland. agentic repl — talks to claude, calls your tools, edits pathways, pushes to prod. all from the terminal. npm i -g bland-cli && bland dev

bland-cli 0.2.11: simulations show the full conversation now. run it, read the transcript, fix the pathway, push, repeat. bland pathway simulate run <id> --wait

npmjs.com/package/bland-cli - Build entire calls from our CLI - Let your model of choice (Claude, Chat, Grok, etc) self simulate and iterate - Publish live - Profit (??)

Things Norm does so you don't have to: - writes the prompt - builds the pathways - adds validation extraction logic - wires up API calls mid-convo - stress tests with agent-on-agent simulations Your personal Forward Deployed Engineer for building voice agents in minutes. See what you can build → hubs.ly/Q048b_Hl0