Products

Products we ship.
Decision infrastructure and AI agents for mid-market operations.

Q: Which product is right for us?

Use the product finder table as a starting point: stockouts and decision lag point to OpsGrid, ERP queries over WhatsApp or Teams point to AskOps, internal document search points to QHub Wiki, and FMCG or export trade operations point to TradeFlow Crew. Most clients start with one product and add a second once the first is live.

Find the right product

If you need to…	Product
Cut stockouts and reduce ops decision lag from days to minutes	OpsGrid ↓
Query your ERP through Teams, WhatsApp, or Slack: no IT ticket	AskOps ↓
Give your team instant search across internal docs and SOPs	QHub ↓
FMCG & Distribution
Run export and FMCG distribution operations end to end	TradeFlow ↓

Built from real deploymentsEvery product here started as a solution to a real customer problem. Open-sourced or productized only after it proved itself in production.

Two licensing tracksCommercial product when it's a vertical category (OpsGrid for Business Central). Open source when it's infrastructure anyone should be able to run.

Self-hosted by defaultYour data. Your server. Your stack. Same posture across both tracks: no vendor lock-in, regardless of how you pay.

· Active beta

Decision Infrastructure for Business Central

Active Beta · Dynamics 365 Business Central · Microsoft Teams

OpsGrid

Signal. Route. Approve. Execute. Audit.

Copilot answers questions. OpsGrid decides and acts.

A purchase order sits in approval with no response clock on any step. A reorder threshold breaches and nobody drafts the PO until someone happens to check. OpsGrid puts a named owner and an SLA on both, in Microsoft Teams, on live Business Central data. Every write still needs your explicit approval. That part never changes.

What is Decision Infrastructure? The definition and the five layers →

Stack: Business Central · Microsoft Teams · Custom AI agents

Model: Signal → Route → Approve → Execute → Audit

Deploy: 3 weeks

Stage: Active beta · 13 spots

See OpsGrid → Apply for beta →

60×

faster decisions2.3 days → 1.2 hours

80%

expedite cost reductionwithin 90 days

BC writeswithout human approval

From one active OpsGrid beta deployment · Dynamics 365 Business Central · mid-market manufacturer · baseline at project start, outcomes at 30 and 90 days.

OpsGrid is built for

Ops Heads & Plant Managers → CIOs & IT Leaders → CEOs & MDs →

· Available · Custom implementation

Conversational ERP for any system

Live · Custom Implementation · Any ERP

AskOps

Conversational ERP interface. Your team sends a message from WhatsApp, Teams, or Slack. AskOps queries your live ERP and responds in seconds. No new app, no screen navigation. Works with Business Central and custom systems today; SAP, NetSuite, and broader Dynamics 365 are on the roadmap.

"Used to call the factory from client meetings to check on an order. Now I check on my phone before I walk in."
Head of Sales, bespoke garment manufacturer

Stack: Any ERP · WhatsApp / Teams / Slack · Custom AI agents

Deploy: 2–4 weeks

Stage: Available · Custom implementation

Interface: WhatsApp · Microsoft Teams · Slack

See AskOps → Book a scoping call →

Open source · Live

Open-source infrastructure

Live · Open Source · Apache 2.0

QHub Wiki

Ask your company's knowledge base from your browser tab: chat interface, instant answers, without opening WikiJS or searching manually. Self-hosted, unlimited users, $0/month. Your data stays on your server.

Stack: WikiJS · Qdrant · Ansible

LLM: Claude / OpenAI / Openrouter

Deploy: ~30-45 min

Users: Unlimited

View on GitHub Managed install: $299 →

Chrome extension: AI answers from any tab

Type a question in the extension and get an instant, AI-synthesised answer from your own knowledge base, without opening a new tab or navigating to the wiki.

Semantic search via Qdrant vector DB

Find the right doc even when the exact words don't match. Qdrant runs locally on your server: no external search API, no data leaving your infrastructure.

Full data sovereignty

Nothing is indexed by third parties. No vendor lock-in. Markdown-first content is portable across any AI platform as the ecosystem evolves.

Any LLM: Claude, OpenAI, or openrouter

Bring your own API key. Claude (Anthropic) recommended for best retrieval quality. Groq available as a free-tier fallback if you need to control costs.

// Powered by QHub · Live in production IntelliconnectQ Bespoke Garment Manufacturing ZapConnectHub Corporate Event Planner

FMCG & Distribution

· Live · FMCG & Export Trade

FMCG Trade Operations Platform

Live · Production

TradeFlow Crew

Order volume outgrows manual coordination before anything else breaks. Buyers call for status. The factory waits on the morning briefing. Procurement gaps surface after the damage is done. TradeFlow Crew closes that gap for FMCG distributors and manufacturers: buyer portal, sales queue, factory queue, rate master with per-buyer overrides, procurement planning. Phase 1 live in production. Phases 2–4 cover supplier chain, export documentation, and overseas buyer management.

Stack: Hono · Cloudflare Workers · SvelteKit · D1 SQLite · Cloudflare Pages

Deploy: 5 days (initial infrastructure)

Stage: Phase 1 live · Phases 2–4 roadmap

Markets: SE Asia · UK · Gulf · North America

See TradeFlow Crew → Book a discovery call →

TradeFlow Crew is built for

FMCG Distributors → Agri Exporters → Enterprises →

New order

Order tracking

Sales queue

Factory queue

In the pipeline

Coming soon

Each tool below is either live or in development. We open-source when it's production-ready. Subscribe for early access.

Supplier Invoice Reconciliation Agent

Q3 2026

A drop-in AI agent that reconciles supplier invoices against purchase orders automatically. Reads invoices, matches to POs, flags exceptions, generates a reconciled report: zero manual intervention. Currently deployed with a manufacturing client reconciling 10 suppliers daily.

Origin: Deployed (manufacturing client, bespoke garment manufacturing). Open-source soon.

Get notified →

Employee Lifecycle Automation Starter Kit

Q3 2026

MS Graph + Power Apps templates for automated employee onboarding and offboarding. License allocation, department head alerts, access removal on departure. Based on a live deployment for an MSP managing 4500+ employees across 7 client companies

Origin: Live deployment (Managed Services Provider, serving 7 client companies).

Get notified →

Tell us your challenge

Bespoke

Have a specific operational problem not covered above? Our products start as client deployments; we may already have something in progress, or we scope and build bespoke. Either way, we size it before we quote.

Start a conversation →

Engineering & Infrastructure

Internal tools. Open sourcing soon.

These Ansible playbooks are used by our engineering team on every client deployment. They handle the repeatable, error-prone parts of server setup (SSL, nginx, Docker, VPN) so we're not rebuilding the same foundation each time. Open sourcing because they're good enough to share.

Ansible: SemaphoreUI Setup

Open sourcing soon

Full SemaphoreUI deployment: nginx reverse proxy, SSL via Let's Encrypt, database setup, and pre-vetted playbook library import. Takes a clean server to production-ready in one run. We use SemaphoreUI as the Ansible orchestration UI across all client infrastructure.

Stack: Ansible · SemaphoreUI · nginx · Let's Encrypt. Linked with the common playbooks library below.

GitHub: coming soon

Ansible: n8n Setup

Open sourcing soon

Self-hosted n8n via Docker, with nginx reverse proxy and SSL provisioning. Takes a clean server to a fully operational n8n instance. The setup takes half a day manually. Used for clients replacing licensed workflow automation platforms.

Stack: Ansible · n8n · Docker · nginx. Phase 2 of our MSP automation deployment runs on this stack.

GitHub: coming soon

Ansible: WireGuard VPN

Open sourcing soon

WireGuard server setup with peer configuration, ufw firewall rules, IP forwarding, and client config generation. How our engineering team accesses client infrastructure securely: faster than OpenVPN, no commercial licensing, repeatable across every engagement.

Stack: Ansible · WireGuard · ufw · Linux. Used on every client infrastructure engagement.

GitHub: coming soon

Common Ansible Playbooks

Open sourcing soon

Internally vetted scripts for the foundation every server deployment needs: server hardening, Docker install, nginx base config, SSL automation, firewall rules, backup configuration, log rotation. Linked with SemaphoreUI so they run from the UI. Not experimental. In active use across all client deployments.

Stack: Ansible · Docker · nginx · SemaphoreUI. Foundation layer for all client infrastructure work.

GitHub: coming soon

Get early access

We notify subscribers when a new tool releases: with install instructions, a demo, and a limited early-access managed install offer. No spam. One email per release.

Your subscription could not be saved. Please try again.

Your subscription has been successful.

Why we open source

We build it. We use it. We share it.

It proves we can ship

A GitHub repo with production deployments is more credible than any case study we could write. Real code, real clients, real infrastructure: visible to anyone who wants to look.

Shared tools create shared understanding

Manufacturers and ops teams that use our free tools know how we think and build before they ever talk to us. For many it's also their first real look at what practical AI infrastructure feels like to deploy and own.

It's the right thing to do

We built on the shoulders of open source tools: Langgraph, WikiJS, Qdrant, Ansible, Clickhouse, Apache Superset, n8n and others. Everything we ship stands on open source foundations. Contributing back is how that ecosystem stays healthy.

Common questions

Frequently asked questions

Which product is right for us?

Use the finder table at the top of this page as a starting point: stockouts and decision lag point to OpsGrid, ERP queries over WhatsApp or Teams point to AskOps, internal document search points to QHub Wiki, and FMCG or export trade operations point to TradeFlow Crew. Most clients start with one product and add a second once the first is live.

What does OpsGrid's "active beta" mean, and what does it cost?

OpsGrid is live in production with a mid-market manufacturer, currently in an active beta with 13 spots open. Contact us for beta pricing. It runs on your existing Dynamics 365 Business Central and Microsoft Teams, live in about 3 weeks.

Is QHub Wiki actually free?

Yes. QHub Wiki is Apache 2.0 licensed, self-hosted, and $0/month with unlimited users. Deploy it yourself in about 30 to 45 minutes, or pay $299 for a managed install. Your data stays on your own server either way.

Do these products connect to our existing ERP, or do we have to migrate anything?

No migration. AskOps is a custom implementation built on an agentic AI framework, answering questions from WhatsApp, Teams, or Slack. Business Central connections are proven today. SAP, NetSuite, broader Dynamics 365, and others are on the roadmap. Each is a fresh implementation scoped to that system's API access. OpsGrid runs directly on your existing Business Central data.

How long does deployment actually take?

QHub Wiki: about 30 to 45 minutes if you self-deploy. AskOps: 2 to 4 weeks for a custom implementation. OpsGrid: about 3 weeks. TradeFlow Crew: 5 days for initial infrastructure, phased rollout after that.

Work with us