.png?width=1920&height=1080&name=Pulse%20Landing%20Page%20Design%20(11).png "Pulse Landing Page Design (11)")
.png?width=100&height=56&name=Pulse%20Landing%20Page%20Design%20(27).png "Pulse Landing Page Design (27)")
Human interpretation for every meeting is expensive and impractical, and it's not ideal for the clerk to manage yet another layer of technology.
Most residents won't ask for an interpreter
In our 2024 National State of Community Engagement survey, 28% of Spanish-speaking residents said they're uncomfortable requesting an interpreter. 71% are comfortable using a translation option that's already available.
Interpreter budgets cover one language during a meeting
Coverage typically pays for one meeting at a time. The Spanish interpreter is booked for the council meeting, leaving the planning commission, the boards and commissions, and the special meeting called on Tuesday go uncovered. Those are where many of the cities biggest land-use and budget decisions are debated.
Another system to maintain
AI tools can solve the problem, but implementing and maintaining them creates another responsibility for some of the busiest people in local government.
More Ways for Residents to Listen In
If your meetings already run on PublicInput, the audio is already flowing in for captions. Interpretation rides the same feed, so there’s no separate audio setup.
.png?width=650&height=366&name=Pulse%20Desktop%20(3).png "Pulse Desktop (3)")
.png?width=778&height=438&name=Pulse%20%26%20AI%20ProjectEmail%20Images%20(7).png "Pulse & AI ProjectEmail Images (7)")
The Resident View
Residents see the exact same public meeting page, just toggled to their language of choice. Captions scroll while interpreted audio plays alongside them.
Web and Phone Use the Same Language Menu
The same set of languages is offered on the page and on the dial-in menu, so a participant who starts on one can switch to the other without learning a different system.
.png?width=650&height=366&name=Pulse%20%26%20AI%20ProjectEmail%20Images%20(10).png "Pulse & AI ProjectEmail Images (10)")
.png "Pulse & AI ProjectEmail Images (11)")
Plug Into Your Existing Tools
If you're already hosting meetings with us, this turns on inside the same setup. No new vendor, login, or hardware is required.
.png?width=160&height=90&name=Pulse%20%26%20AI%20ProjectEmail%20Images%20(12).png "Pulse & AI ProjectEmail Images (12)")
Voice-Matched Audio
Audio is built on Azure Cognitive Voice, preserving the speaker's gender and tone, so their voice doesn’t get lost.
.png?width=160&height=90&name=Pulse%20%26%20AI%20ProjectEmail%20Images%20(13).png "Pulse & AI ProjectEmail Images (13)")
One Flat Annual Number
We use your actual meeting schedule to set the scope and give you one annual price per board, covering every language you need. If meetings run over or you need an emergency session, you’re still covered.
.png?width=158&height=89&name=Pulse%20%26%20AI%20ProjectEmail%20Images%20(14).png "Pulse & AI ProjectEmail Images (14)")
Integrated Into Your Boardroom
Join the meeting through your existing public meeting page. In-room captions match your branding as well, making it easy for everyone to follow along.
We Don't Sell Hours
Most AI Interpretation is sold as a bank of hours you buy upfront. We price per board, flat, for the year. Here is the difference in practice.
.png?width=320&height=180&name=Pulse%20Landing%20Page%20Design%20(25).png "Pulse Landing Page Design (25)")
Live Meeting Interpretation runs on Microsoft Azure, keeping your data in the U.S.
Everything You Need to Know About Live Meeting Interpretation
Does this replace our interpreters?
No. It lets you spend your interpreter budget where it does the most good.
Staffing two interpreters through an entire four-hour meeting is hard, expensive work, and it only covers one language. Running routine coverage on the technology frees that budget for the places a person is better: legal proceedings, high-stakes hearings, and ASL, which this product does not touch at all.
Some agencies do move routine meetings off human interpretation. Others keep every interpreter they have and add the languages they are unable to staff. Our read is that there's plenty of need to go around, and your interpreters are working where their skill is needed.
How accurate is it, really?
The engine underneath is Azure AI Speech. Microsoft Research reached "human parity" on the standard conversational speech benchmark: the AI got 5.8% of words wrong, and professional human transcribers got 5.9% wrong on the same test. The production models are improving.
Captions and the recording make the errors visible, so a mistake can be caught and corrected. Spoken interpretation just disappears. The underlying models keep improving with feedback, and because we build on the current state of the art, you get the latest models as they ship.
Could we just use YouTube auto-translate or Zoom captions?
For a stream, sometimes. For a public meeting there are gaps worth knowing about. YouTube does not auto-translate captions on a live broadcast, only on the replay. Google Meet's translation does not run on live streams at all, by Google's own documentation.
None of them ring a phone. There is no keypad language menu, no per-language audio, nothing on a display in your chamber, and the captions are not part of your meeting record. A resident without a smartphone gets nothing from any of them.
Whether you stream through YouTube, eScribe, or Granicus, we give you a way to show live captions in up to 130 languages.
Does this satisfy our language access obligations?
That is what this is built for.
Most state and federal language access requirements come down to two things: real-time access to the meeting in the languages your community speaks, and translated access to the material around it. We cover both. Live interpretation and captions during the meeting in every language you enable, and translation of your online content, including agendas, notices, and meeting notes.
SB 707 is the current California example, and it explicitly allows machine translation. Title VI and locally adopted language access policies point in the same direction.
You also get a record of what was provided at each meeting, which is usually the hardest part to produce when someone asks. Tell us which requirements you are working against and we will show you how the coverage maps to them. We are committed to keeping you compliant with state and federal requirements.
What do you need from our AV setup?
An audio feed we can reach. If your meetings already run through PublicInput, we have it. If there's an encoder or a streaming appliance in the chain, we'll get on a call with that vendor and work out the handoff before you commit to anything.
Often it's simple. A laptop on the audio output, feeding the stream, is enough.
Where does the audio go, and is anything stored?
Translation and voice run on Azure Cognitive Services inside our Azure environment, exclusively on U.S. servers.
The meeting recording and captions are securely stored for retrieval after the meeting, and you can provide access to them for residents online if you choose.
AI models are not trained on your data, and you retain full ownership and control of your data.