# Every AI you’ve built is *blind.* Yours *doesn’t* have to be

AI shouldn’t need another prompt to know something changed

Presence ambiently notices confusion, absence, attention changes, and shared visual context, then responds without interrupting work.

Try Presence free [Talk to the founder](https://presence.bohita.com/#book)

90 seconds · camera on · Google sign-in

Image: A stylized woman working quietly at a table in a warm room

She paused.
Looked away.
Sighed.
**It noticed**

You’ve seen it notice.

[Talk to Bohita](https://presence.bohita.com/#book)

Try it free

Product trials and onboarding

## Sit with everyone who tries your product

Your best prospects get a founder watching them install, stepping in at the right moment, and learning what made the product difficult. Presence makes that available to everyone.

[See how product trials work](https://presence.bohita.com/try-your-product)

After the Session

01 **Where they got stuck**

02 **What it helped with**

03 **What it left behind**

## It sees the person, not just the screen

Notices when someone goes quiet, looks away, or gets stuck. Then it speaks without being asked.

## It keeps talking while it thinks

Slow tools and deep reasoning run in the background. No dead air.

## Long conversations don't get more expensive

Minute 60 costs the same as minute 1. Meetings, debugging, onboarding, and support can keep going.

Ambient interaction

## The prompt is not the only channel

Presence notices what people don’t put into prompts.
It knows when to help, wait, or keep working.

[How Presence works](https://presence.bohita.com/#method)

You look confused **It offers help**

You step away **It stops expecting a response**

You return **It knows you are back**

You share something **It can react to what you are working on**

Privacy, plainly

## It sees enough to notice

Presence reads coarse signals such as present, looking away, screen shared, and engaged. It does not send a continuous video feed.

When the reasoning model needs to look, it receives a low-resolution frame. Higher resolution is available when the task calls for it. Frames are not retained in later conversation history.

The launch demo is not recorded. Other sessions require separate, session-specific consent before recording.

[Read security and privacy](https://presence.bohita.com/security)

You are building…

Choose one to see the setup.

### Onboarding or support

It steps in when someone gets stuck, then reports where they got stuck, what it helped with, and what it left behind.

You’d write “Help when someone gets stuck. Do not wait to be asked.”

What comes back

Where they got stuck, what Presence helped with, and what it left behind.

[See the workflow](https://presence.bohita.com/try-your-product)

### An AI product that needs a face

Drop a character into your app that sees your user and reacts.

You’d write “You are attentive, concise, and present.”

```
POST /v1/sessions
{
  "external_id": "your-user-id",
  "surface": { "type": "web" },
  "startup": "managed"
}
```

[Build this](https://presence.bohita.com/#integrate)

### Something that joins meetings

Same agent in Teams, Meet, or Zoom.

You’d write “Follow the meeting. Speak only when you can move it forward.”

What you provide

A Teams, Meet, or Zoom meeting URL. Presence joins with the same prompt and tools.

[Build this](https://presence.bohita.com/#integrate)

### Something that calls people

Approved outbound calls with the same agent.

You’d write “Keep the call warm, brief, and useful.”

What you configure

Choose an approved outbound call path, then choose who Presence should call. Inbound calls remain carrier-originated.

[Build this](https://presence.bohita.com/#integrate)

One URL. One iframe

## Put Presence in your product

[Get your API key](https://presence.bohita.com/account/api-keys) Buy 60 minutes for $9

WEB · THE WHOLE CLIENT INTEGRATION Copy

```
<iframe src={session.embed_url} allow="camera; microphone; autoplay; display-capture" />
```

[Read the installation guide](https://presence.bohita.com/llms.txt)

How it works

## The parts are nameable. Keeping them together is the work

Four engineering decisions make the difference between a stack of models and a live conversation that holds together. [See every measurement](https://presence.bohita.com/benchmarks)

01 / TURN STATE

### The face follows the conversation

Turn state, not audio amplitude, decides where it looks and when it shows processing. Mouth articulation is derived from the actual output audio, so it cannot drift from speech.

**0.65 ms *median* 0.83 ms *p95*** Expression inference on CPU. 25 ms causal analysis window, 10 ms recurrent updates.

02 / PLAYBACK CLOCK

### Audio is authoritative

Expression frames travel immediately before the PCM packet they describe, addressed in 16 kHz samples. If rendering stalls, stale poses are dropped instead of replayed late.

**~155 ms** Source canvas to decoded WebRTC at 30 fps.

03 / ASYNC WORK

### Tools do not stall the turn

The model dispatches work and keeps talking. Results arrive when the conversation is next idle; the video above shows the longer call returning while the session continues.

**22.1 s *and* 9.7 s** Two concurrent think calls in the recorded demo session. Neither blocked a conversational turn.

04 / SPECULATION

### Silence is spent generating

Presence starts generating at a provisional endpoint while it keeps listening. Resumed speech discards the work. Confirmed silence releases audio already made.

**1.8–2.1 s *p50* ~1.2 s *speculative*** Warm full turn, microphone speech to first audio. A native-live baseline was faster on the same fixture: 1.054 s versus 1.955 s.

WHAT AN HOUR COSTS

## Most voice APIs charge you more for the same conversation as it goes on

They keep the conversation in memory and bill you for all of it, every turn. Turn one is cheap. Turn forty re-sends everything that came before it.

**Presence bills occupied time, not accumulated history**

An hour-long meeting costs what an hour costs. The same is true for debugging, onboarding, or support. The rate stays flat.

**17×** What turn 40 cost versus turn 1 on a hosted realtime API over 40 turns, for the same amount of talking. [Read the forty-turn benchmark.](https://presence.bohita.com/posts/what-happens-on-turn-forty)

This is why hour-long sessions are viable here.

OBSERVATION STREAM RECORDED · 01:36

```
{
  "event": "pause",
  "signals": {
    "gaze": "away",
    "silence_ms": 2840
  },
  "note": "possible hesitation detected"
}
```

PRICING PER OCCUPIED MINUTE

**Free** Five 5-minute sessions. Google sign-in. No card.

**$0.15** $9 for 60 minutes. Buy 60 minutes

**$0.10** $100 for 1,000 minutes. Buy 1,000 minutes

Occupied time only. Credits never expire.
