Skip to content
Data Forge

We build the platform. Then we run it.

Data Forge designs, ships and operates production software for Australian businesses: the services, the data, the AI agents, the infrastructure underneath and the measurement on top. We run our own group of companies on it first.

Estate snapshot, measured from the repository on 2026-09-09

Running services, by plane

2026-09-09

Status Service Role Runtime
Platform & infrastructure
Deployed auth-engine Rust
Deployed auth-idp Rust
Deployed authz-projector Go
Deployed config-vault Rust
Deployed telemetry-gateway Go
Data platform
Deployed intel-event-pipeline Go
Deployed intel-clickhouse-consumer Go
Deployed intel-orchestrator Python
Deployed graph-sync Go
Deployed analytics-etl Go
Applied AI & agents
Deployed agent-runtime Go
Deployed knowledge-graph Go
Deployed forge-context Go
Deployed code-intel-mcp Go
Deployed rag-memory-mcp Go
Deployed embedding-gateway Go
Product engineering
Deployed core-api Go
Deployed surebi-api Go
Deployed surebi-reports Go
Deployed solar-api Go
Deployed quote-calculator Rust
Deployed finance-api Go
Growth & measurement
Deployed intel-collector-api Go
Deployed intel-lead-scoring Go
Deployed intel-capi-dispatcher Go
Deployed intel-compliance Go
Deployed growth-studio-api Go
Automation for operators
Deployed notification-handler Go
Deployed workflow-engine Rust
Deployed scheduling-optimizer Rust
Deployed lead-assignment-worker Rust
Deployed realtime-gateway Go

Showing 32 of 79 services and workers in the estate.

See the whole board

Six planes. One estate. All of it operated by the people who built it.

Most consultancies have a services list. This is a running system. Each band below is a plane of that system, what it promises, and the real services that carry it today. Measured from the repository on 2026-09-09.

Running services, wired to their stores

2026-09-09

54 connections declared in the deploy manifests.

32 nodes · 6 stores

Platform & infrastructure

The layer nothing above it should have to think about.

Kubernetes on immutable, API-managed nodes. GitOps reconciliation, one image lane with provenance gates, and a release train so a deploy is a decision rather than an accident. Identity, secrets, certificates, storage and backups are part of the build, not a follow-up ticket.

Full detail
  • Talos Linux Kubernetes, nine nodes in Perth
  • Argo CD GitOps with pinned manifests and a release train
  • One image lane with provenance and entrypoint gates
  • OIDC identity: auth-engine sessions, auth-idp provider, SpiceDB
  • config-vault secrets with provenance and audit

In the estate

  • auth-engine
  • auth-idp
  • authz-projector
  • code-intel-toolchain
  • config-vault
  • config-vault-cli
  • config-vault-reconciler
  • +1 more
The k8-forge fleet, drawn to the hardware One Dell PowerEdge R740xd 2U server above a 4 by 2 grid of eight HP EliteDesk 800 minis, each with a status LED. U2 U1 node-9 Dell PowerEdge R740xd node-1 6c / 16 GB node-2 6c / 16 GB node-3 6c / 16 GB node-4 6c / 16 GB node-5 12c / 32 GB node-6 12c / 32 GB node-7 12c / 32 GB node-8 12c / 32 GB
Nine machines, drawn to the hardware facts; arrangement schematic.

Data platform

One store, one job. Nothing quietly doing three.

Most data problems are placement problems: events in the transactional database, relationships in a JSON column, embeddings in whatever was nearest. We put each kind of data in the store built for it and keep every derived layer rebuildable from its source.

Full detail
  • Postgres as the transactional system of record
  • ClickHouse for append-only events and analytics, queried in SQL
  • Redpanda as the transport between services, with declared topics and dead-letter queues
  • FalkorDB graphs for identity and code relationships, queried in Cypher
  • Qdrant vector search for embeddings, tenancy enforced as a payload filter

In the estate

  • analytics-etl
  • analytics-service
  • data-pipeline
  • dataforge-pg-migrations
  • graph-sync
  • intel-clickhouse-consumer
  • intel-event-pipeline
  • +3 more

Applied AI & agents

Agents wired into real systems, not a chat box on the side.

We build agent runtimes and tool servers that read and write the actual estate: the work ledger, the code graph, the event bus, the warehouse. Retrieval runs over your own corpus with tenancy and consent already enforced, so an agent can be useful without being dangerous. Nothing gets write access before it has been evaluated.

Full detail
  • Agent runtime and orchestrator services
  • MCP tool servers over Postgres, ClickHouse, Redpanda, Qdrant, Prometheus and OpenProject
  • A code-intelligence graph for impact analysis before a change ships
  • Knowledge graphs for entities and relationships
  • An embedding gateway that routes to Gemini or a local Ollama model

In the estate

  • agent-api
  • agent-kb-embedder
  • agent-kb-mcp
  • agent-orchestrator
  • agent-runtime
  • clickhouse-mcp
  • code-intel-mcp
  • +16 more

Product engineering

The right language for the job, in one coherent codebase.

Go for domain services, Rust where the work is CPU-bound, Python where the work is data and machine learning, typed React on the front and Astro for the sites. One monorepo, one build graph, and shared packages for the things every surface needs, so a capability is built once and used everywhere.

Full detail
  • Go domain APIs on a standard service contract
  • Rust workers for quoting, scheduling, search and media
  • Python services for documents, reports and property intelligence
  • Vite and React operator consoles and customer apps
  • Astro marketing sites built for Core Web Vitals

In the estate

  • cashflow-api
  • core-api
  • deal-checklist
  • document-processor
  • finance-api
  • image-processor
  • integrations-api
  • +12 more

Growth & measurement

Numbers you can defend when the spend is questioned.

Client-side analytics quietly lies. Consent gating, ad blockers and platform attribution all bend the shape of the truth. We collect server-side, score the bots out, deduplicate conversions into Meta and Google, export raw events to the warehouse, and run ad accounts against real jobs rather than platform-reported conversions.

Full detail
  • Server-side event collection with browser fingerprint capture
  • Bot scoring before an event counts
  • An identity graph and recency-weighted lead scoring per brand
  • Meta and Google Conversion APIs with event deduplication
  • GA4 with raw export to BigQuery

In the estate

  • cms-analytics-sync
  • growth-os-api
  • growth-studio-api
  • intel-activation
  • intel-analysers
  • intel-analysers
  • intel-capi-dispatcher
  • +5 more

Automation for operators

The repetitive work handled, permanently.

Lead routing, notifications with a delivery record per lead, franchise and territory workflows, scheduling, accounting and CRM integrations, and internal consoles so an owner can see the whole business without asking anyone. This is the original promise of Data Forge: technology that gives people their time back.

Full detail
  • A durable notification lane with a delivery record per lead
  • Round-robin lead assignment and territory routing
  • A workflow engine for work-item transitions and triggers
  • Schedule optimisation for field availability and conflicts
  • Xero and Pipedrive integrations

In the estate

  • command-centre-api
  • csv-import-worker
  • lead-assignment-worker
  • notification-handler
  • realtime-gateway
  • scheduling-optimizer
  • workflow-engine

The estate, counted.

Every claim on this site maps to something in the repository or the cluster. These are the numbers a script reads out of the codebase, not a marketing estimate.

Repository

Services and workers
79 (55 Go, 16 Rust, 8 Python)
Apps and sites
14
Shared packages
9
Kubernetes manifests
496
SQL migrations
609
CI workflows
37
Commits since 2025-12
7,390
Merged pull requests
2,937

Cluster

k8-forge · 9 nodes

One Dell PowerEdge R740xd (72 cores, 256 GB, an NVIDIA P40) and eight HP EliteDesk minis running Talos Linux, in Perth, reached over Tailscale.

Stores, one job each

Postgres
system of record
SQL
ClickHouse
events and analytics
SQL, columnar
Redpanda
transport between services
Kafka API
FalkorDB
relationship graphs
Cypher
Qdrant
embeddings
vector search
MinIO
objects and backups
S3 API

All six stores run on node-9 by design: the one machine in the fleet built to hold state.

Nine nodes, one job each

node-1 HP EliteDesk 800 G5 6c / 16 GB control plane
node-2 HP EliteDesk 800 G5 6c / 16 GB control plane
node-3 HP EliteDesk 800 G5 6c / 16 GB control plane
node-4 HP EliteDesk 800 G5 6c / 16 GB worker
node-5 HP EliteDesk 800 G4 12c / 32 GB worker
node-6 HP EliteDesk 800 G4 12c / 32 GB worker
node-7 HP EliteDesk 800 G4 12c / 32 GB worker
node-8 HP EliteDesk 800 G4 12c / 32 GB worker
node-9 Dell PowerEdge R740xd 72c / 256 GB, NVIDIA P40 data plane23 workloads pinned

Runs the SURE group first.

Data Forge operates a group of Australian service brands on the same platform it builds for clients. Every system on this page carries real bookings, real leads and real invoices before it is offered to anyone else.

  1. SURE Building Inspections

    surebuildinginspections.au

    Building and timber-pest inspections across Western Australia and New South Wales, run as franchise territories.

    Systems

    • A twelve-thousand-page Astro site generated from a content pipeline, with a searchable index of roughly ten thousand building defects
    • Lead capture, booking and franchise routing through a Go API
    • Server-side tracking, conversion APIs and the Google Ads operation behind them
    • Inspection-report intake, parsing and a defect knowledge base
  2. SURE Renewables

    surerenewables.au

    Residential solar and battery systems.

    Systems

    • A solar CRM with quoting, proposals and installation pipeline
    • A Rust quote calculator for pricing rules and margins
    • Schedule optimisation for installer availability
    • Xero integration and customer notifications
  3. SURE Building Services

    surebuildingservices.au

    Building maintenance and repairs.

    Systems

    • A customer-acquisition site with a lead lane straight into the core API
    • Quoting and job management on the shared platform
  4. SURE Pest Control

    On the shared platform

    Pest management, on the same platform primitives configured for a different service line.

    Systems

    • Marketing site and enquiry flows
    • Shared lead routing and notifications

Also built for clients

  • Beauty

    A booking and business-management platform with online payments, an admin dashboard and its own booking and customer services.

  • Financial education

    A client intake wizard and an adviser cockpit that turns a household budget into consequences a couple can feel, built for live sessions on a big screen.

  • Hospitality

    A specialty coffee brand site and the API behind its events and products.

How work moves.

The same flow runs our own estate and every client engagement. Four of the six stages end at a gate you hold. Nothing ships unverified, and merging is never the finish line.

  1. Scope

    gate · path

    We write down the goal, the success criteria, the systems touched and the people needed. You see it before anything else happens.

  2. Design

    gate · spec

    One-way doors, meaning data models, wire contracts and store choices, are designed and agreed on paper. Reversible things ship lean.

  3. Plan

    gate · plan

    The spec becomes an ordered task list with a file-ownership map, so parallel work never collides.

  4. Build

    no gate

    Specialists build in isolated worktrees against a shared ledger. You can see what is in flight and what is blocked without asking.

  5. Verify

    no gate

    Tests, lint and type gates across everything affected, then adversarial review for anything high-stakes. Red does not ship.

  6. Ship

    gate · ship

    Merging deploys nothing. An image is built, pinned, released and reconciled, and the change is done when a probe shows it running.

What done means

A change is done when the pod is running the new image and a probe shows traffic. Not merged, not staged, not "should work".

Who is behind it.

Build things that keep working when nobody is watching them.

Jacob, founder of Data Forge. An advanced degree in AI, machine learning and data science, applied to business. Taught data science at Curtin University and the Institute of Data. More than ten years running companies, including a marketing business with over fifty staff. Owner-operator of the SURE group.

I used to run a multi-million dollar marketing company with over fifty staff. On paper it looked like success. Behind the scenes I was constantly putting out fires, working around the clock, and slowly burning out.

Then everything changed. I lost my mum, and I realised how much I had missed while buried in the grind. I left the company, moved home, and began rebuilding with a new goal: create freedom, not stress.

The whole story, and how it shaped the way we work

Open a conversation.

Tell us what you run today and what keeps breaking. We will tell you, plainly, what we would do about it and whether we are the right people to do it.