Hey Product Hunt! 👋

I'm Marcin, a network security engineer with multiple CCIEs and more firewall migrations behind me than I care to count.

Every migration followed the same painful pattern: export the config, open a massive spreadsheet, spend two weeks manually mapping objects and rewriting rules — then pray nothing breaks during the maintenance window. The existing tools are either deprecated (RIP Palo Alto Expedition), vendor-locked, or hidden behind a "call us for a quote" wall.

I got tired of the grind and built NetConverter.ai.

🧠 What makes this different from "just using AI"

This is the question I knew would come up, so let me address it directly.

NetConverter is not an LLM wrapper that pastes your config into a prompt and hopes for the best. The translation engine is built on a PatternDB we've developed over years of real-world migrations — a structured knowledge base that maps every meaningful conversion path between vendors: object types, service definitions, rule semantics, NAT behaviors, and application identification across L4–L7.

Most tools stop at L4. We map App-ID and application-layer policies to their equivalents across vendors — something no general-purpose AI tool has and no spreadsheet process captures reliably.

AI in our pipeline executes workflows against that knowledge base — it handles ambiguity resolution, edge case classification, and validation. It doesn't hallucinate translations. The PatternDB does the heavy lifting; AI handles what's genuinely uncertain.

The result: 95% translation accuracy, validated across 44 production migrations.

🔒 Enterprise and security-conscious teams

We know your firewall config is crown jewel data. You're not pasting it into ChatGPT.

For enterprise deployments, NetConverter installs locally on your network — your configs never leave your environment. You get full API access to our translation and PatternDB engine, or if you're running an AI agent workflow, MCP (Model Context Protocol) access so your agents can query our knowledge base directly without any config leaving your perimeter.

This is the path for regulated industries, federal environments, and anyone under DLP or compliance constraints that rule out cloud AI tools entirely.

🚀 What ships today

Multi-vendor parsing: Cisco ASA, FortiGate, Palo Alto — with Juniper SRX, Check Point, and others in the pipeline
Human-in-the-loop review: automated mappings with full engineer sign-off before anything is finalized
Import-ready output: clean configs for FMC, Panorama, and FortiManager
Pre-flight audits: shadowed rules, object conflicts, unsupported features flagged before your maintenance window
Behavioral equivalence checking: we prove permit/deny equivalence on sampled traffic flows — not just "it looks right," but verified correct

🎯 Who this is for

Network security engineers, MSPs, and consultants running multi-vendor refreshes — especially those left stranded by the Expedition sunset. If you've spent a weekend in a spreadsheet mapping 3,000 address objects, this is for you.

💰 Try it free

Free tier available for small configs and evaluations. Paid plans cover production workloads. Enterprise plans include local deployment, API access, and dedicated support.

Interested in the beta program or a live demo? Visit netconverter.ai — I'll run your config through the engine today and show you the side-by-side results versus what a manual process produces.

💬 Three things I'd love your feedback on

The Pain: What part of your last migration burned the most hours?

The Roadmap: Juniper SRX, Check Point, SonicWall, or Meraki — which should I prioritize?

The Site: Anything unclear on the landing page? I'm fixing things in real-time today.

I'll be here all day. Happy to do a live demo or answer anything technical.

Thanks for checking it out. 🙏

Marcin — Founder, NetConverter.ai