Product How It Works Developers Roadmap Investors Docs/GitHub Get Started
Governance Infrastructure for Agentic AI

AI Agents That Know When to Stop

Haltio pauses high-risk agent actions — refunds, deletes, sends, deploys — for human approval before they ship. Open source. 2-minute integration.

Cross Agent Governance  ·  2-minute integration  
🐍 Python SDK ⚡ 2-min Setup 💬 Slack Alerts
MVP · Building in public since 2026
RefundAgent · checkpoint #4471
process_refund(order_id="8841", amount=$412.00)
HIGH RISK
Approve
Block
Illustrative preview — dashboard UI in active development.
Why Haltio Exists

Hi I'm Devesh, I spent my life savings building an ecommerce store "Kawaii404.com". To scale it, I invested heavily in AI agents — automating support, refunds, and order handling, as I was confident in AI as a Computer Science Student.

One day, I discovered the agent had refunded 80% of the store's profit back to customers, triggered by a single misread policy edge case. Nobody had approved it. Nobody even saw it happen until the damage was done.

Agents should move fast. But the actions that can actually hurt a business need a human checkpoint first.

Unchecked agent actions

Agents executing irreversible actions with no review step before they happen.

Approval fatigue

Too many low-value interruptions train humans to rubber-stamp everything.

Enterprise trust gap

Pilots stall because teams don't trust agents enough to actually deploy them.

Concurrent agent chaos

When 20 agents hit a decision point at once, who decides first — and what depends on what?

"Enterprise AI pilots are not failing because agents are incapable. They are failing because teams don't trust them enough to deploy."

0
Developers in active beta
0
Integration time
AI
Licensed & open source
0
Frameworks supported
0
Human-reviewable high-risk actions
How It Works

Every High-Risk Action Gets a Human Checkpoint

A risk classifier sits between your agent's decision and its execution. Operators (the developers or compliance leads who configure agent permissions) define what's safe to automate — and what isn't.

Agent Task Risk Classifier LOW RISK auto-proceed (within scope) MEDIUM RISK queue for review HIGH RISK block + notify Human Decision Approve → Continue Block → Halt + Log

LOW-risk tasks proceed automatically only if they fall within the agent's configured permission scope, defined by the operator at setup.

Timeout, escalation path, and fallback behavior are configurable per agent and per task type. Example: timeout_seconds=300, on_timeout="escalate"

The Real Moat

Built for Multi-Agent Systems.

When 50 agents hit a checkpoint simultaneously, Haltio queues, prioritizes, and routes approvals without blocking your pipeline. Dependency-aware. Timeout-safe. Rollback-ready.

Priority Queue

High-stakes checkpoints surface first when many agents request approval at once.

Dependency Graph

Tracks which agent actions depend on which approvals, so nothing executes out of order.

Configurable Timeout & Escalation

Set a timeout per task type, with a defined fallback if no human responds in time.

Non-Blocking Pipeline

One agent waiting on a checkpoint doesn't freeze the others.

Fail-Safe / Fail-Open Modes

Choose what happens if Haltio itself goes down — block everything, or proceed with full logging.

Slack-Native Alerts

Approve or block directly from Slack — no context-switching to a separate tool.

For Developers

Built by Developers, For Developers.

One decorator. Any agent framework. No architecture changes.

checkpoint_example.py
from haltio import checkpoint

@checkpoint(risk="high", on_timeout="escalate", timeout_seconds=300)
def process_refund(order_id, amount):
    ...
LangChain
AutoGen
CrewAI
LlamaIndex
Custom agents

Haltio wraps any agent's action-execution layer with a single decorator. No framework lock-in, no architecture changes.

Production Architecture (in development)
Agent SDK Message Queue (Redis / SQS) Approval Orchestration Service Dashboard UI Slack Audit Store

Is data encrypted?

Checkpoint payloads are encrypted in transit (TLS 1.3) and at rest.

How long is data retained?

No task data is stored by Haltio beyond a configurable retention window.

Can we self-host?

Yes — self-hosted deployment is available for data-residency requirements.

What if Haltio goes down?

Configurable fail-safe (block all pending actions) or fail-open (proceed with full logging). Enterprise tier targets 99.9% SLA with multi-region redundancy.

Why not just use LangGraph or the OpenAI Agents SDK?

You can — for a single agent on a single framework. LangGraph's interrupt() and the OpenAI Agents SDK's native approval flow both work well there. Haltio is for the moment you have agents built on more than one framework, or more than one agent hitting checkpoints at the same time — that's a vendor-neutral, cross-framework problem no single framework owner is incentivized to solve for its competitors' agents too.

Honest status

What you see here is a working, production-ready MVP — a real open-source SDK, not a mockup. It's Phase 1 of a much larger system: a full multi-agent orchestration and enterprise compliance platform. The roadmap below shows where this is going. We're looking for developers to help build it.

Contribute on GitHub
See real commit history — no curated highlight reel, just the log
Roadmap

MVP Today. Platform Tomorrow.

Three phases, in order, with no skipped steps.

Phase 1

Now

Shipped

Open-source Python SDK with human-in-the-loop checkpoints.

Phase 2

Multi-Agent Orchestration

~6 months

Concurrent approval graphs, SLA management, escalation logic.

Phase 3

Enterprise Compliance Platform

12–18 months

RBAC, SSO, audit trails, EU AI Act reporting.

For Investors

We're Building the Compliance Layer Enterprises Will Be Required to Have.

Regulatory tailwind

The EU AI Act mandates human oversight for high-risk AI systems. Enterprise security teams increasingly require audit trails before approving any agentic deployment. This is a tailwind, not a headwind, for a product like Haltio.

Market framing

Figures vary by report scope. Research and Markets projects the AI governance and compliance tooling market at roughly $0.6B in 2026, growing to roughly $2.6B by 2030 at a ~44% CAGR. Other analyst firms scope the market differently and land higher — we're citing the conservative, named estimate rather than averaging conflicting reports into a number nobody can trace.

ICP — Ideal Customer Profile

The wedge

AI agents that touch money — refunds, payouts, order changes — inside teams shipping their first production agent.

Series A–C B2B SaaS companies with a dedicated ML/AI engineering team (5–50 engineers), deploying LLM-based agents into production workflows for the first time, where a single agent mistake has direct customer or financial impact.

Draft wedge, not yet confirmed against the beta cohort — narrow this to wherever the 5+ beta developers are actually concentrated (e.g. a specific vertical) once that's known.

Customer discovery

Place reserved for customer testimonials

Reserved for quantified testimonials:

Business model

Like Sentry, we start with a free open-source SDK developers add in minutes. The enterprise layer — audit trails, RBAC, compliance dashboards — is where revenue lives. The SDK is the top of funnel; the platform is the business.

Competitive landscape

ToolWhat it actually is
LangGraph interrupt()Single-agent, synchronous only, no dashboard
OpenAI Agents SDKNative pause/resume approval flow, but scoped to OpenAI's own SDK — no cross-framework or cross-vendor layer
HumanLayerFramework-agnostic approval routing over Slack/email — single-agent, no concurrent dependency graph
Prefect / TemporalWorkflow orchestration, not agent-specific, no Slack-native UX
Build-it-yourselfMonths of engineering, no standard protocol
HaltioMulti-agent concurrent, SDK + dashboard, 2-minute integration

We're not first to ship human-in-the-loop approvals — LangGraph, the OpenAI Agents SDK, and HumanLayer all support it today, and we'd rather say that plainly than have you find it yourself. What none of them do is manage approval dependencies across many concurrent agents running on different frameworks in the same pipeline. That's the gap Haltio fills, and why a framework or model vendor isn't the one positioned to build it — a vendor-neutral audit layer has to work the same way across competitors' agents too, which isn't something any single framework owner is incentivized to ship.

Why this is hard to copy

approvals.json plus a webhook is a weekend build for a single agent — we know, because that's roughly what the simplest version looks like. What's harder to replicate is the layer above it: a dependency graph tracking which actions depend on which approvals across many concurrent agents, configurable timeout and escalation logic per task type, and — on the roadmap — a policy engine with fine-grained, per-action permissioning that makes Haltio the one integration point across multiple agent frameworks inside an org. That's sustained orchestration and reliability engineering, not a weekend script.

Pricing model

Free

Unlimited agents, community support.

Pro

Per-workspace monthly fee. Team dashboard, Slack integration, audit logs.

Enterprise

Custom pricing. RBAC, SSO, compliance reporting, SLA.

Use of funds

01

Engineering — build the production multi-agent orchestration backend.

02

Developer growth — open-source community, docs, integrations.

03

First enterprise pilot — dedicated customer success.

The ask

Raise amount to be finalized directly with Devesh — we'd rather have that conversation by email than publish a placeholder number on this page.

12 months from this raise

TODO — fill in once the ask is set

Reserved for the milestone story tied to the ask: e.g. "Raising $X gets us to Y design partners, Z paying enterprise pilots, and a shipped policy engine (Phase 2) within 12 months." Partners are funding what the raise buys in a year, not the product as it exists today — this is where that gets stated explicitly.

We have a working SDK, 5+ developers already testing it, and a team that has lived this exact problem firsthand. If you invest in developer infrastructure, let's talk — email .

Team

One Founder. One Real Problem.

D

Devesh

Founder, Haltio

Building governance infrastructure for AI agents after watching firsthand what happens when nobody's watching: his uncle's ecommerce store lost 80% of its profit to a single unapproved agent action. That's not a hypothetical case study — it happened to family, which is why Devesh is building the fix now, while agent adoption is outpacing the governance layer underneath it, instead of waiting for someone else to.

LinkedIn - Click here to connect with me
TODO — state plainly

Commitment level isn't stated yet. "Full-time since [month/year]" reads very differently from a side project to a partner deciding whether to back the person, not just the idea — fill this in directly rather than leaving it implied.

5+

Developers currently in active beta.

TODO — highest-leverage item on this page

Reserved for one named design partner: company name, logo, and one quote attributed to a real person Devesh can stand behind. A single named company — even a 10-person startup, even unpaid — carries more weight here than any amount of polished copy.

Get Started

Add a Human Checkpoint to Your Agents in 2 Minutes.

Email Devesh