Vero agents · Early access

You’ve built agents.
Now run them.

Not a framework. A runtime — session management, message routing, tool execution, memory, and agent-to-agent communication. Already wired.

One message spawns a full team
A single trigger creates a session, plans tasks, and launches parallel agents — no glue code.
DC
David Cohen
I need help with my denied claim, policy #IL-4821
C
Coordinator
On it. Let me assemble the right team for this…
plan_tasks → decompose into 3 parallel lanes
subscribe ×3 → register brain callbacks
T
Transcript
real-time STT
P
Policy
CRM + knowledge
S
Sentiment
emotion + compliance
3 agents active · 3 session lanes open · coordinator waiting for callbacks
01 · Positioning

CrewAI gives you agents. LangGraph gives you a graph. Neither gives you infrastructure.

CrewAI and LangGraph are agent frameworks. They solve the agent loop — the part where an LLM decides what to do next. That’s real, and it matters.

But it’s about 10% of what you need to run agents in production.

The other 90% is everything around the loop: session management so two runs don’t collide. Message routing so agent replies actually reach the user. Tool execution that composes without a sprawl of conditionals. Memory that persists across conversations. VM isolation so a rogue tool call can’t take down the system. Real-time event delivery to your frontend. Cron scheduling so agents can run on a timer. Auth, tenancy, observability.

That’s not a framework problem. That’s an infrastructure problem. And it’s what you end up building yourself — with Redis, Postgres, a WebSocket server, a homegrown task queue, and a lot of duct tape — because no framework provides it.

Vero Agents isn’t a better framework. It’s a different thing: a runtime. The infrastructure your agents need to actually run, not just think.

02 · How it works

Trigger. Execute. Coordinate.

01

Trigger

A message arrives — from a user, a webhook, a cron job, or another agent. The runtime creates a session, resolves the agent's workspace, and enqueues the run on a FIFO session lane. No concurrent collisions. No manual locking.

02

Execute

The agent enters a multi-turn LLM loop (max 10 turns by default). Each turn: stream from Claude via custom SSE parser, execute tool calls through a decorator chain, append results, loop. Tools compose — brain, messaging, CRM, code execution, MCP integrations, device control — without conditionals. Static chain at startup, per-run handlers added based on what the agent can do.

03

Coordinate

Agents don't chat with each other. They write findings to a shared brain — a structured memory service with semantic, episodic, and graph storage. Other agents subscribe to scopes. When data appears, the brain pushes a notification. The coordinator doesn't poll. It registers a callback and goes idle until the brain wakes it. Zero agent-to-agent messages.

03 · Telepathy

This is what “zero direct messages” actually looks like.

Walk through a real orchestration: an insurance call with 4 agents. Watch how they coordinate entirely through shared brain state.

Agent actions
Insurance call · 4 agents coordinating through shared brain state.
  1. 1
    C
    Coordinator
    plan_tasks + subscribe ×3
    Coordinator plans parallel lanes and registers brain callbacks — then goes idle.
  2. 2
    T
    Transcript
    write_brain(transcript)
    Transcript agent writes a Hebrew STT chunk to shared brain state.
  3. 3
    B
    Brain
    notify → Sentiment
    TELEPATHY
    Brain subscription fires. Sentiment agent is notified — no one sent it a message.
  4. 4
    S
    Sentiment
    read_brain(transcript)
    Sentiment reads the transcript — automatically, through the brain.
  5. 5
    P
    Policy
    write_brain(history)
    Policy agent writes CRM customer history into shared state.
  6. 6
    B
    Brain
    notify → Sentiment
    TELEPATHY
    Second brain callback fires. Sentiment cross-references transcript + history.
  7. 7
    S
    Sentiment
    write_brain(alert)
    ALERT
    Sentiment writes an alert: frustration 0.82 · escalation 0.91.
  8. 8
    B
    Brain
    callback → Coordinator
    DECISION
    Brain wakes the Coordinator. It reads full state and decides to escalate to a human.
Brain state
Empty — waiting for agents to write.
0 direct messages exchanged
Step 0 of 8
“The Transcript agent didn’t send the Sentiment agent a message. It wrote to the brain. The brain notified the subscriber. Two agents coordinated without a single chat message, webhook, or API call between them. That’s telepathy.”
04 · Sandcastle

Your agents need a sandbox.
Not your production server.

When an agent runs untrusted code — web scraping, file manipulation, arbitrary tool calls — it shouldn’t run on your infrastructure. It should run in an isolated, disposable VM that boots in 125ms and dies when the agent is done.

Sandcastle is Vero’s execution environment for agent workloads. Secure microVMs. On-demand. Per-agent isolation. Three image tiers (base, browserbase, dev-machine), each with progressively more capabilities. Secrets injected at boot via MMDS — never written to disk.

~125 ms
cold boot time
for a microVM
12 tools
MCP server
ships with every sandbox (filesystem, browser, code execution, screenshots…)
0
shared state
each VM completely isolated. A rogue agent can't reach another agent's data.
Sandcastles
test
20:40
Files
~/workspaceclick files to view
Terminal
System Monitor
runningtest
CPU

5%

4 vCPUs

Memory

219 / 3943 MB

6% used

Disk

1.3G / 3.9G

34% used

Uptime

0m

Processes

84

Running processes

PID

1077122

CPU: 4c MEM: 4096MB
Container
Agent
MCP
What’s inside every sandbox
Same image contract for every VM. No custom bootstrapping per agent.
  • MCP Server
    12 tools, running on :3000. Any MCP-compatible agent framework can connect.
  • REST API
    on :8080. Trigger tools, read results, stream events.
  • Headless Chromium
    screenshot, accessibility tree, JS execution, navigation.
  • LiveKit streaming
    audio capture for voice agent workloads.
  • Secret injection
    MMDS metadata service at boot. API keys, tokens, credentials — injected, never persisted.
05 · Architecture

What’s under the hood.

Session lanes

FIFO queues per agent/conversation/sender. No concurrent runs, no manual locking, no race conditions. Idle cleanup built in. Each lane is independent — one slow agent doesn't block another.

Decorator-based tool chains

Tools compose through a decorator pattern: brain → messaging → CRM → code → MCP → cron → spawn. Static chain at startup; per-run handlers added based on agent capabilities. No if-else routing. Add a new tool = add a decorator. Every tool degrades gracefully when unconfigured.

Three-tier execution

Same agent definition, three environments — In-process (conversational, lowest latency), Sandcastle microVM (isolated, untrusted tools), Anthropic-managed (hosted execution via Anthropic). The runtime routes based on config. Switch tiers by changing a value, not rewriting code.

Real-time message routing

Postgres for state, ClickHouse for analytics, Redis Streams for real-time events. WebSocket delivery to frontends. Agent replies route back through the messaging service to all conversation participants. Not bolted on — the messaging layer IS the coordination fabric.

06 · Compare

This isn’t a framework comparison. It’s a category difference.

CrewAI and LangGraph are agent frameworks — they solve the LLM loop. Vero Agents is an agent runtime — it solves everything around it.

CrewAILangGraphVero Agents
What it isAgent frameworkStateful graph frameworkAgent runtime
LLM loopYesYesYes
Session managementNo — you build itNo — you build itFIFO lanes with idle cleanup
Message routingNoNoFull messaging service (Postgres + ClickHouse + Redis)
Tool compositionManual wiringManual wiringDecorator chain, composable at startup
Inter-agent coordinationAgents chat with each otherState machine transitionsBrain telepathy — shared state, zero messages
Isolated executionNoNoSandcastle microVMs (~125ms boot)
Persistent memoryBring your ownBring your ownBuilt-in brain (semantic, episodic, graph, tasks)
Real-time frontend deliveryNoNoWebSocket + SSE + Redis Streams
Cron schedulingNoNoBuilt-in poll-based scheduler
Agent-to-agent with governanceNoNoAllowlist-routed A2A
Auth & multi-tenancyNoNoJWT + bearer token, row-level security
CrewAI and LangGraph are good at what they do. But what they do is 10% of running agents in production. The other 90% is infrastructure — and that’s what Vero is.
07 · Developer experience

Trigger an agent in one call.

# Trigger an agent run via gRPC
grpcurl -d '{
  "agent_id": "coordinator",
  "conversation_id": "call_8291",
  "sender_id": "david-cohen",
  "message": "I need help with my denied claim"
}' localhost:50051 agent.AgentService/TriggerRun

gRPC for agent execution. REST for messaging and admin. WebSocket for real-time delivery. SSE for event streams. MCP for sandbox tools. All authenticated, all tenant-scoped.

08 · Use cases

Built for agents that work together.

Customer service teams

An inbound call triggers a coordinator. It spawns a transcript agent, a policy agent, and a sentiment agent — all in parallel session lanes. They coordinate through the brain. The coordinator sleeps until a callback wakes it. Decision made in seconds.

Autonomous research teams

A research question triggers a planner. It spawns specialists — search, analysis, writing, review — each in a Sandcastle VM with isolated browser access. Findings flow to the brain. The planner synthesizes when all lanes complete.

Internal operations agents

Cron-scheduled agents that run daily — inventory, reports, compliance. Each agent has its own workspace, tool chain, and session lane. Results persist in the brain for downstream agents or human review.

09 · What you don’t have to build

Everything you’d spend 6 months building. Already done.

  • Session management that doesn't race
    FIFO session lanes
  • Message routing between agents and users
    Messaging service with Redis Streams
  • Tool execution that doesn't explode
    Decorator chain with graceful degradation
  • Memory that persists across conversations
    Brain service (semantic, episodic, graph, tasks)
  • Isolation for untrusted agent tools
    Sandcastle microVMs
  • Real-time updates to your UI
    WebSocket + SSE delivery
  • Cron-triggered agent runs
    Built-in scheduler
  • Agent-to-agent communication with governance
    Allowlist-routed A2A
  • Multi-tenant auth
    JWT + bearer token, row-level security
  • Observability
    Agent telemetry, run logs, event history
10 · Early access

The waitlist is open.

Vero Agents is in early access. If you’re running multi-agent systems in production — or about to — we want to work with you. Waitlist members get:

  • API access as soon as your account is provisioned
  • A month of Pro-tier usage included
  • Direct line to the engineering team
  • First look at SDKs, MCP gateway, and self-hosted deployment

Your agents deserve infrastructure.
Not another framework.