FEATURES

Everything your agent team can do

A desktop-native platform for voice-controlled AI agents that see your screen, coordinate across machines, and never lose context.

VOICE-FIRST INTERFACE

Talk to your agent team like you talk to your people.

Wake word activation, natural language commands, and real-time two-way voice powered by OpenAI's Realtime API. Ask questions, trigger tasks, share your screen — no keyboard required.

VOICE INTERFACE

YOU

“Hey Yma, draft the Q2 report...”

YMA

“On it. Pulling last month's data...”

REALTIME VOICE · MULTIMODAL SCREEN SHARING

"Hey Yma" wake word — on-device ONNX, no API key, works offline

Real-time two-way conversation via OpenAI Realtime API

In-context questions without switching windows

Voice-triggered task execution and delegation

Multimodal screen sharing — Yma sees what you see

AGENT ROLES

RESEARCH

PLANNING

SCHEDULE

DOCUMENT

WRITE

MEMORY

RECALL

COORDINATING ACROSS EVERY MACHINE

MULTI-AGENT ARCHITECTURE

Four local specialists. Four remote agents. One voice command.

Four local agents — Research, Planning, Document, Memory — plus four remote coding agents via SSH delegation (Claude Code, Codex, Gemini CLI, OpenCode). They plan, delegate, and execute across your machines — no handoff management from you.

RESEARCH AGENT

Searches the web, pulls structured data, and synthesises findings from any source.

PLANNING AGENT

Breaks tasks down, assigns priorities, and builds schedules your team can execute against.

DOCUMENT AGENT

Drafts, edits, and formats reports, briefs, and long-form documents on command.

MEMORY AGENT

Retains context across every session. Recalls decisions, preferences, and patterns without being asked.

FLEET ORCHESTRATION

Your agents run 24/7 — across every machine you own.

Yma Agent deploys across workstations, servers, and cloud VMs, forming a private mesh of AI agents that coordinate and execute without supervision.

FLEET TOPOLOGY

OFFICE

SERVER

LAPTOP

TEAM

ENCRYPTED MESH VIA TAILSCALE

MULTI-MACHINE

Deploy across your infrastructure

Install on workstations, office servers, cloud VMs — even your team's laptops. Each instance is a node in your private AI network.

AGENT DELEGATION

Agents talk to each other

The Research agent on your office machine can hand findings to the Planning agent on your laptop. Context and knowledge sync automatically.

AUTONOMOUS SCHEDULING

Agents plan ahead

They create, schedule, and assign tasks to each other in the Planner workspace. Your AI team runs while you sleep.

ENCRYPTED MESH

Zero-trust Tailscale network

Every connection is encrypted end-to-end over Tailscale. No open ports. No cloud relay. A private mesh VPN only your devices can access.

CROSS-MACHINE MEMORY

Shared context everywhere

Memory, context files, and knowledge sync across every machine in your fleet. No cold starts on any device.

OPERATIONAL INFRASTRUCTURE

The foundation that makes everything else possible.

Memory that persists. A harness that extends. Automation that runs while you focus.

PERSISTENT MEMORY

Memory that persists across sessions and machines. Topic-based knowledge files, team preference learning, and a full decision history — your agents always know the context.

▪Decision history and audit trail
▪No cold starts — ever

OPERATIONAL HARNESS

A complete operational harness — shared project and business knowledge, persistent facts and session history, and custom skill plugins tailored to your workflows.

▪Tailscale VPN for secure remote operations
▪Convex-based task tracking with fleet-wide sync

AUTOMATION ENGINE

Scheduled task execution — morning, midday, evening. Agents work in the background while you focus on what matters.

▪Automated backups and sync verification
▪Custom workflow triggers

REACH & OUTPUT

Generate anything. Deliver it anywhere.

CONTENT GENERATION

AI IMAGE GENERATION

AI image generation from voice descriptions

VIDEO PIPELINES

Video generation via Sora 2 Pro, Veo 3.1, and Kling 2.5

PEN DISPLAY

Pen display annotation — sketch edits, AI interprets

DOCUMENT TEMPLATES

Document templates and report generation

MARKETING ASSETS

Marketing asset creation from a single command

CROSS-PLATFORM MESSAGING

Telegram bot with native long-polling and forum topic routing

DISCORD

Discord gateway bot with multi-machine channel routing

WhatsApp via Twilio with HMAC-verified N8N webhooks

SLACK

Slack Events API with HMAC-verified N8N webhooks

BIDIRECTIONAL

Send and receive — bidirectional communication

FILE SHARING

File sharing across all platforms

One command. An entire team in motion.

See how Yma Agent deploys across your infrastructure and works while you focus on what matters.

BOOK A DEMO