Nicheloom

Market intelligence for builders — see what's gaining traction before it's crowded.

AxonFlow, governing LLM and agent workflows

Details

External ID
46692499
Source
HN
Company
—
Product
—
Website domain
—
Launched
Jan. 20, 2026
Cohort
—
Upvotes
11
Upvotes percentile
0.5335968379446641
Tags
—
Fetched at
Sept. 7, 2026, 9:25 p.m.
Updated at
Sept. 7, 2026, 9:25 p.m.

Description

Hi HN, we’re building AxonFlow for teams running LLMs or agents in real production systems.Once agent workflows move past demos, failures are rarely model issues. They tend to show up as execution problems during real runs.Short 2-minute technical demo showing execution control and auditability in practice: https://youtu.be/FNgnESo9RtIAxonFlow is a self-hosted, source-available (BSL 1.1) control plane that sits inline in the execution path and governs LLM calls, tool calls, retries, approvals, and policy enforcement step by step. It does not replace your orchestrator and can run alongside LangChain, CrewAI, or custom systems.The problems we focus on are usually discovered only after going to production: - retries that accidentally repeat side effects - partial failures mid-workflow - permissions that differ per step - limited ability to inspect or intervene during executionThis is not aimed at early demos or hobby projects. It’s for teams already operating under real production constraints.GitHub: https://github.com/getaxonflow/axonflowDocs: https://docs.getaxonflow.comI’d value feedback from folks running LLM or agent workflows in production.

Enrichment

Theme
AI agent frameworks and developer tools
Vertical
Horizontal
Function
Workflow automation
Audience
Developer
AI stance
AI-native
Project type
Commercial product
Normalized one-liner
governance for llm and agent workflows
Manually corrected
False

Could you build this?

No AxonFlow is an enterprise-grade runtime governance and execution control plane for agent workflows. Building deterministic guardrails, policy enforcement proxies, and auditability at low latency requires specialized distributed systems and security engineering.

What it would actually take: The architecture comprises an ultra-low-latency reverse proxy (Go or Rust) intercepting LLM/tool calls, evaluating deterministic policy engines (e.g., OPA/Rego), enforcing execution state transitions, and writing immutable audit logs. The primary difficulty is inline policy enforcement and state-machine validation without adding latency or breaking streaming connections. It requires deep expertise in systems programming, enterprise security, and proxy architecture.

Discussion

14 comments analyzed.

Competitors mentioned: LangChain, LangGraph, CrewAI, Kaos (kaosThere)

Concerns raised: Migration cliff from gateway to proxy mode without rewriting orchestration code, Policy complexity becomes bottleneck beyond simple PII/SQLi rules, Tail latency (p95, p99) with overlapping policies and cold cache not benchmarked, Risk of becoming orchestration layer competing with LangChain/CrewAI, Latency overhead relative to non-LLM operations like tool calls

Feature requests: Image redaction support, Deterministic replay for stateful workflows, Policy simulation and staged rollout tooling, Performance benchmarks with complex rules and cold cache scenarios

Competitors

Other products that read as similar to this one — 257 launches clear the similarity bar, closest 8 shown.

Attention rank: #132 of 258 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).

Launched 80 days after the earliest competitor.

Other launches for this product

Same idea, different domain

Nobody's really built a workflow automation tool for Real estate yet.