One graph. Every function. Answerable.
Mid-market businesses run on a dozen disconnected tools and a few overworked people holding it all together in their heads. Daxo replaces the swivel-chair and the spreadsheet glue with a single intelligence layer: the knowledge graph as the connective tissue, and an AI you can simply talk to as the interface to all of it.
Every quote, invoice, email, supplier, contract, and conversation becomes a node in a per-tenant graph — the Daxo Atlas.
"Which invoices are overdue?" · "What did we actually decide on the Henderson matter?" · "Who still hasn't paid us?"
You ask — and the answer comes back in seconds, straight from the live business. No chasing three people, no spreadsheet that's a week out of date. The answer you'd normally lose half a day to is one question away.
The platform is modular: tenants adopt the functions they need — Sales, Customers, Suppliers, the AI email workspace, Finance, Legal, an executive Command Center, and more — each a productised vertical, each feeding the same graph. Eleven of thirteen modules are already deployed and serving in production, running on the same multi-tenant, security-gated chassis.
You didn't build your business to hand its data to someone else's cloud. With Daxo, the data, the access, and the intelligence stay yours. Your records live in a database that is yours alone — protected so that any request without your verified identity returns nothing: it fails shut, not open. The AI reads across your whole business without your documents ever leaving our infrastructure to be understood, and every answer runs on your own AI key — so you own the model relationship, and CORAZ never bears its cost, retains what you ask, or resells it. We operate the platform; we are not the custodian of your content.
We're candid about the journey: deepening this into formal, independently-audited controls — tenant-authorised break-glass access, SOC 2 — is part of what this raise funds. The architecture is built for it; this round matures it.
The mid-market is glued together by hand.
Fragmented systems
A typical A$5–50M business runs accounting, CRM, email, payroll, and documents in separate tools that don't talk. The "integration" is a person copying between them.
Knowledge trapped in people
The state of the business lives in one or two key staff. When they're away — or leave — the answers leave with them. Nothing is queryable.
Clerical labour at scale
Invoice entry, email triage, order tracking, chasing receivables, assembling reports — hours of low-leverage work that grows linearly with the business.
Incumbent SaaS deepens the problem: each tool adds another dashboard to check and another silo to reconcile. The owner who wanted leverage ends up managing software. Daxo is built for the buyer the brand was designed around — owners running multiple businesses, portfolio managers, and the office managers who actually operate the business.
An AI-native operating system.
Daxo inverts the model. Instead of bolting AI onto a dashboard, the graph comes first and every module writes into it.
One surface over every system
Sales, email, suppliers, finance, legal, people — every function reports into one place. Nothing hides in a silo only one staff member knows how to open.
Every action is a node
Modules mirror each business record into the tenant graph in the same database transaction — if the write rolls back, so does the graph. The Atlas is never stale and never a separate sync job.
Ask Daxo, in plain language
A shared "Ask Daxo" interface sits over the whole graph — semantic search, entity and audit queries, synthesis. The same engine powers staff questions and the AI that drafts, classifies, and reconciles.
And it's modular and isolated: each module is a deployable unit with its own database and fail-closed row-level security. A tenant buys only what they need, and the data boundary is enforced by the database kernel, not by convention.
Thirteen modules, one chassis.
Each module is a complete business function with a clear advantage, running on a shared multi-tenant chassis (own database, fail-closed row-level security, entitlement + security gates). Every module both removes manual labour and writes into the graph for real-time answers.
Status verified 22 Oct 2026 against each module's production /healthz on its Fly host, plus a signed operator read of the prod Cortex control plane. Hover any chip on a card to see how that benefit is delivered.
Live in production · 11 of 13 modules deployed & serving
Sales
LiveFull quote-to-cash: opportunity pipeline → manufacturing/order management → invoicing → receivables, with two-way Xero/MYOB reconciliation. Vertical-aware (consulting, manufacturing, financial planning).
Customers (CRM)
LiveUniversal relationship registry — customers, suppliers, contacts, activities — with a smart-intake wizard, AI field-validation (ABN checksum, DNS, consistency), and an inbound-email touchpoint inbox.
Suppliers
LiveBuy-side mirror of the CRM: vendor registry (risk, terms, lead-time, MOQ), purchase orders → goods receipts → one accounts-payable bill per receipt, pushed to Xero through the platform hub.
Daxo Workspace
LiveInstitutional-grade email intelligence: ingest → classify → route → reconcile → file. Google Workspace & Gmail are connected and live — mail is received, classified, routed to the right module, reconciled against invoices, and attachments extracted automatically. AI drafts replies; a human always sends.
Finance
LiveBookkeeper cockpit with emailed-bill intake: extract, human-review, code, and push supplier bills to Xero. A human reviews before any push; every bill carries payment status and aging, so "what's overdue to pay?" is answerable live.
Legal OS
LiveA legal-practice operating system: matter dashboard, Bates-numbered evidence, evidence-anchored chronology, correspondence triage, a doctrine knowledge base, and per-/cross-matter AI chat with hybrid search — all on self-hosted embeddings (no document text leaves the platform).
Command Center
LiveThe executive control plane: a strategy → performance → projects cascade. A ratified strategy becomes the tenant's North Star that other modules align to; it binds to live module KPIs and answers the executive graph in plain language.
Daxo UX surface
Live · platformThe product face: AI-first landing, module launcher, context-aware drawer, actionable search, and graph/timeline visualisation — plus an operator-only "Almanac" (the company's book of record: Chronicle, Trends, Diagnostics).
An org of agents — the business, run by a team you command.
Every function in Daxo has a named AI agent watching it, and one agent above them all — Tomo, the Overseer — synthesising the whole picture. This isn't a chatbot in a sidebar; it's an organisation chart of intelligence, each agent wired to the live graph of the function it owns. You stop reading dashboards and start running a leadership team: glance across your agents, and the one with its hand up is the only one that needs you.
Status you can trust
Each agent's state — Thriving, Steady, Strained, Critical, Standing by — is derived from real metrics (cash runway, pipeline coverage, PO match rate, inbox SLAs, churn), never hand-set. A red badge means the numbers are red.
A CEO's-eye view
Tomo reads across every function and writes the one-line brief a founder actually wants — what's on plan, what's flagged, and which agents are waiting on a data source. One glance replaces ten tabs.
It escalates before it breaks
Agents don't wait to be asked. Gen stops the line when supply breaks; Sora raises the alarm before a quarter is lost; Ryo blocks a payment until a person clears it. Attention is routed to you, early.
The specialists · one named agent per function, bound to its module and the apps that feed it
Hoshi
Command CenterStrategy & cadence. Holds the whole-business picture — plan tracking, performance, open risks — and calls it out when a department slips or the plan drifts.
Sora
Market IntelligenceCommercial. Pipeline coverage, win rate and slippage — an honest forecast, and an early alarm when coverage can't make the number.
Gen
Core OperationsThe floor, where the work is. PO match rate, deliveries and exceptions — and it stops the line for a human the moment supply breaks.
Correo
Strategic IntelligenceInbox & comms. Reads everything that arrives, routes it, and flags what's ageing before an SLA breaches — your Gmail, already connected.
Ryo
GovernanceFinance & risk. Runway, receivables and compliance flags — and the hard stop: nothing is paid while a fraud or risk hold is open.
Kiku
Workforce IntelligencePeople & workforce. Stands by until you link a people, recruitment or safety app — then activates automatically. An honest empty state, not a broken one.
Ryo watches the money and the risk register together. While runway is long, receivables are clean and no holds are open, Ryo stays quiet and thriving. The instant a fraud hold opens or runway drops below your threshold, Ryo turns Critical — exactly as shown above — surfaces the offending items, and freezes the payment until a named person approves it. The agent never pays; it prepares, flags, and waits for human authority. That is operational safety you can point an auditor at.
Every agent is customisable per tenant: tune what "strained" and "critical" mean for your business, choose which apps feed each agent, and adjust its brief — all over a platform-versioned definition, so you inherit improvements without losing your changes. Two guarantees make it safe to hand to staff: a narrow-never-widen merge law means a customisation can only tighten the security floor, never widen it; and per-staff projection shows each person only the briefs and graph reads their role entitles — enforced at a runtime tool broker, not in the UI. Customise freely; the floor holds.
Live · building out · deployed on the live chassis, domain depth maturing
People
Live · buildingWorkforce intelligence: the org-structure primitive other modules consume, plus roster, compensation, entitlements and leave — with tiered PII authorisation and money kept out of the graph.
Core Operations
Live · buildingThe operations function group — methodology, knowledge, partnerships — reading peer modules over signed channels rather than duplicating their data.
Client Comms
Live · buildingClient-facing documents, onboarding workflows, e-signature on engagement agreements, and info-gathering checklists — all graph-tracked.
Portfolio Mgmt
Live · buildingTenant portfolio and investment-research platform: append-only transaction ledger, document import with semantic retrieval, and a living per-ticker financial model. Validated against real research data.
In build
Marketing — Daxo Presenter
In buildDescribe a video in plain words; Daxo writes the script and renders a branded AI-presenter video, delivered as a finished file — broadcast-quality marketing without a studio. GPU render proven end-to-end; sellable add-on next.
Timesheet
Built · launch parkedA Cortex-native timesheet: work sessions flow in automatically, attribute to clients, and surface an "AI-leverage ratio" headline. Built and multi-tenant hardened; production launch intentionally parked behind higher-priority modules.
Cortex is the platform engine tenants never see: the knowledge-graph store, the "Ask Daxo" retrieval, and a metered AI gateway. Tenant AI usage runs on the tenant's own key through a seat/quota/model-policy gate — so the platform meters AI without bearing the cost of inference, and embeddings are self-hosted on-platform (no document text leaves to a third party). This is both the moat and the margin story.
Live-in-production states verified 22 Oct 2026 by probing each module's production /healthz endpoint on its Fly host and a signed operator read of the prod Cortex control plane. "Building out" modules are deployed and serving on the live chassis with their domain feature-set still maturing; Timesheet and Presenter are not yet deployed.
Built safe, cheap to run, secure, and live.
Four properties an institutional buyer underwrites — engineered into the platform, not bolted on. Each is enforced in the running system today, with the honest frontier flagged.
Operational safety
- Human-gated AI. The system drafts, classifies, and reconciles — but never sends mail, pays a bill, or creates a ledger entry on its own. Every AI write is an auditable card with undo.
- Guardrails on the autonomous builder. Write-scope ceilings and dangerous-command interception are enforced by a commit-time hook; a two-strike denial rule stops and escalates rather than retrying blindly.
- Tamper-evident audit log. Every action appends to a log protected against update/delete, with a cryptographic hash-chain over operator actions.
- A release can't reach a customer without passing a real-browser, fresh-tenant acceptance run and a security gate (below).
Wage minimisation
- Clerical work removed, not relocated. Email triage, invoice matching, order tracking, CRM entry, chronology assembly, time attribution — each module targets a role a growing business would otherwise hire.
- One analyst for the whole business. "Ask Daxo" answers across every module, replacing the staff member who assembles reports by hand.
- Leverage, measured. The Timesheet module surfaces an AI-leverage ratio — the automation story made quantitative per tenant.
- Scales sub-linearly. The business grows; the clerical headcount doesn't have to.
Security
- Database-kernel tenant isolation. Postgres FORCE row-level security on every tenant table; the runtime role cannot bypass it; unset tenant context raises rather than leaking. The AI layer inherits the same boundary.
- A quantified launch gate. A 43-test adversarial security matrix (mapped to OWASP ASVS, OWASP LLM Top 10, NIST) with 31 hard gates; a module cannot become sellable without a GREEN verdict — proven by a negative control.
- Signed service identity. Cross-module calls use per-issuer EdDSA keys with rotation; misconfiguration fails closed.
- Frontier: least-privilege service authz and a kernel sandbox are specified and reviewed, not yet shipped; two compliance gates (GDPR erasure, restore drill) activate at first prod customer.
Real-time information
- The graph is the source of truth. Every business action is a node, written in the same transaction as the business record — operational truth and audit trail are the same substrate.
- Answerable in plain English. Semantic, entity, and temporal queries over the whole business, served live over a streaming API.
- Self-reconciling. A periodic integrity scan keeps the graph convergent with the underlying systems; drift is repaired automatically.
- Lawful-purpose data governance is already enforced in production across the live tenant's modules.
The "talk to your business" product line.
The near-term roadmap clusters around one thesis — a business you can talk to — expressed as chat, video, mobile, and AI document review, with a parallel security-hardening track readying the platform for paying multi-tenant customers.
| Initiative | What it adds | Investor relevance | Status |
|---|---|---|---|
| Daxo Chat | A multilingual team-chat module where @daxo answers from the graph; on-the-fly EN/ES/VI translation delivers each message in the recipient's language. | Sticky daily-use product; pulls multilingual SME staff off Slack/WhatsApp, feeding the graph. | Queued |
| Daxo Presenter | AI-presenter marketing video from a plain-language brief, rendered on rented GPU. | Sellable premium add-on; GPU render already proven end-to-end. Closest to shippable. | P3 done |
| Daxo Mobile | React Native capture app — speak or type on the go, durable offline, syncing into the graph. | The premium consumer face of the workspace; native-app beachhead. | Queued |
| Email intelligence (Workspace & Gmail) | Mail received, classified, routed, reconciled and attachments extracted — already live in the Daxo Workspace. Per-seat Gmail OAuth extends it to every staff mailbox. | The headline labour-saver, delivered — proof the roadmap converts to shipped product. | Live |
| AI document review | Real metered AI review inside the docs surface. | Turns document QA from stub to billable. | Reviewed |
| Security hardening track | Least-privilege service authz, cryptographic non-repudiation, command-safety unification. | Readies the platform for live multi-tenant customers and enterprise diligence. | In flight |
Invoice #INV-2043 to Henderson Pty Ltd is 7 days overdue — A$18,400, due 15 Oct. The goods receipt has cleared; it's waiting on finance approval.
Modular subscriptions, high-margin AI.
Predictable recurring revenue from modules, usage revenue on top, and an AI cost structure that doesn't scale against us.
Most "AI SaaS" pays for every model call its customers make — inference is a cost of goods that grows with usage and erodes the margin. Daxo doesn't. Tenant AI runs on the tenant's own key through our metered gateway — we meter it, seat-gate it and cap it, but we don't carry the inference bill. Embeddings are self-hosted on our own hardware. The result is software-grade gross margins that hold as usage grows.
Three tiers that expand with the customer
Start on the graph
The Daxo Atlas, Ask Daxo, and one or two modules — typically Sales or the Workspace. The wedge: one function, immediate leverage.
Run the whole business
The full module suite — Sales, Customers, Suppliers, Finance, Workspace, People — one graph across every function.
Multiply yourself
One operator, several businesses, each on its own Daxo, consolidated into a single portfolio seat — monetising the portfolio-operator segment directly.
Large all-module tenants reach A$8–15K+/mo.
Pricing combines a base platform fee with per-active-user and AI-usage components — mirroring how the comparable stack (CRM, AI copilots, vertical tools) already charges per seat, and capturing the value a flat per-org fee would leave behind.
Two revenue engines
- Recurring (predictable ARR): the tier subscription plus per-seat add-ons where usage is personal — staff mailboxes, mobile capture.
- Usage (expands with adoption): AI overage above the cap, and premium outputs like Daxo Presenter videos, priced per use.
Land and expand
A tenant lands on one module and expands across the graph as adjacent functions switch on — every module added lifts revenue inside the same account. Net revenue retention is the growth engine, targeting 108% NRR.
Daxo is not a prototype dressed up for a raise. It is eleven independently-deployed production services, each with its own isolated database, already wired to the systems a real business runs on — two-way Xero reconciliation and live Google Workspace mail ingestion among them. The graph beneath them is live and accumulating real records, not a seeded demo.
On 22 October 2026 every one of those services answered a production health check, and the control plane answered a signed operator read — independently verifiable, not asserted. The entire platform runs on roughly A$800 a month of cloud and SaaS infrastructure — Fly, Cloudflare, Google and Xero combined: the capital-efficient, metered architecture isn't a future promise, it is already how the platform operates. A seed round here does not fund "build the product" — the product exists and runs. It funds putting it in front of customers.
Metric targets are deliberately conservative and benchmark-grounded (SaaS Capital 2025, for a A$25–50K-ACV mid-market SaaS): gross margin sits below the 75–80% steady-state band in the early years; NRR of 108% is inside the 102% median → 111% top-quartile range; LTV/CAC is held at the 3:1 "healthy" line. Blended ARPA (~A$42K) sits inside the researched tier pricing. The metered-AI structure is the primary gross-margin lever.
The mid-market, Australia first.
Who buys
Australian SMEs with A$5–50M turnover, and owner-operators running multiple businesses or portfolios. The office manager recommends; the principal approves. Daxo is sold on force-multiplication, not features.
The portfolio operator
A fast-growing segment: one entrepreneur running several businesses, each doing real turnover. Put each business on its own Daxo and the owner runs the whole portfolio from one seat — every entity visible, every entity answerable, with no management layer to staff inside each one. It is the brand promise delivered literally: multiply yourself, across every business you own.
Why Daxo, not SaaS
Point tools (Xero, HubSpot, Slack) each solve one slice and add a silo. Daxo sits above them as the connective graph — it integrates the ledger rather than replacing it (Xero stays the book of record), which removes switching friction and makes the business queryable as a whole.
| Approach | What they give | The gap Daxo fills |
|---|---|---|
| Horizontal SaaS (Xero, HubSpot, Monday) | A dashboard per function | No connective layer; integration is manual labour |
| Point AI copilots | Chat bolted onto one app | No cross-system graph; can't answer the whole business |
| Bespoke consulting builds | Custom, one-off | Not productised, not multi-tenant, not repeatable |
| Daxo | One graph, every function, plain-language | The second-in-command layer — productised and multi-tenant |
What a mid-market firm pays today
A representative ~50-person firm assembling best-of-breed tools — published list prices, AUD / month:
| Function | Typical tool | A$/mo |
|---|---|---|
| Accounting & payroll | Xero Ultimate | ~180 |
| AP automation | BILL · 4 seats | ~400 |
| CRM & sales | HubSpot Sales Pro · 15 seats | ~2,100 |
| AI copilots | Microsoft 365 Copilot · 50 seats | ~2,350 |
| Disconnected stack — subtotal | four vendors, four silos, no shared graph | ~A$5,000 |
| + Legal practice vertical | Clio / LEAP · seats | +2,800–5,000 |
| Daxo — one connected platform | Business tier · every function on one graph, Ask Daxo included | A$4,000–7,000 |
Daxo lands at or below the cost of the tools it replaces — and adds the one thing none of them have: a single knowledge graph the whole business can be asked in plain language. For legal-vertical firms the saving is larger still.
Illustrative projection.
A conservative bottom-up path, grounded in the pricing, market and benchmark research — not a placeholder.
Revenue trajectory · illustrative
Assumptions behind the curve
- Tenants: 12 → 38 → 95 → 180 → 285 — founder-led in Y1–2, inside-sales ramp from Y3; growth decelerates each year (~220% → 60%).
- Blended ARPA ~A$42K / yr (A$3.5K/mo) — within the researched tier bands and SaaS Capital's A$25–50K mid-market ACV range.
- Gross margin 60% → 78% over Y1–Y5, after the metered-AI pass-through (benchmark steady state 75–80%).
- NRR 100% → 108%; GRR 85% → 90% (SaaS Capital 2025, A$25–50K ACV band).
- Crosses the ~A$1–2.5M Series-A ARR bar in Year 2; Year-5 ARR is still <0.5% of TAM.
Sources (full URLs in the research pack): business counts — ABS Counts of Australian Businesses (cat. 8165.0, Jun 2026); cloud/SaaS market — Gartner AU public-cloud 2025; retention, margin & efficiency benchmarks — SaaS Capital 2025, Bessemer, KeyBanc; pricing comparables — published vendor pricing (Xero, HubSpot, Salesforce, Microsoft Dynamics 365, Clio, Smokeball, LEAP). Projection figures are illustrative and founder-owned.
Raising A$1M to reach live revenue.
The platform is built; eleven modules are in production. This A$1.0M seed funds the step from engineered to live, revenue-generating SaaS — go-to-market, the security, compliance and legal work that paying customers and diligence require, the first full-time hire, and the founder's capacity to finish and launch.
Use of funds · A$1.0M · 80% build / launch / legal / first hire · 20% founder (indicative split)
| Area | Allocation | Outcome |
|---|---|---|
| Go-to-market & first customers | A$180k · 18% | First paying tenants; a repeatable sales motion |
| Engineering & product delivery | A$220k · 22% | Chat, Presenter and Mobile to general availability |
| First full-time hire | A$140k · 14% | Delivery / GTM capacity beyond the founder |
| Security, compliance & legal | A$160k · 16% | Close security gates; independent testing & verification; company, contract, terms and IP paperwork |
| Infrastructure & deployment | A$100k · 10% | Reliable per-tenant deployment at scale |
| Founder (full-time capacity) | A$200k · 20% | Runway to complete the build and launch — see §11 |
Milestones this round buys
- [__] paying tenants across [__] verticals
- Production launch of Daxo Chat and Presenter
- All launch security gates GREEN; compliance gates exercised
- A$[__] ARR and a defined path to Series A
Every decision filters through one test: does it move Daxo closer to being a live, revenue-generating SaaS? Infrastructure that doesn't unblock launch waits.
- Months 0–3
First paying tenants
Founder-led conversion of the live platform into revenue; security gates GREEN.
- Months 3–6
Daxo Chat to GA
Multilingual team chat live; the first hire onboarded.
- Months 6–9
Presenter + inside sales
Premium add-on live; a repeatable, non-founder sales motion.
- Months 9–12
~A$1M ARR
Crosses the Series-A ARR bar with NRR above 100%.
- Months 12–18
Series-A ready
Proven motion and a defined path to the next round.
The founder.
An operator who understands how a business runs, a disciplined real-time decision-maker, and the person who has already built the platform.
Daniel Lopez is an accountant by training who built and sold his own accounting practice before moving full-time into Forex trading — a business built the hard way, over a long and uneven road, and his primary income for years since. Two habits were formed there and run through everything after: understanding how a business works from the books up, and making decisions under uncertainty on live, real-time information.
For the past three years Daniel has worked full-time in software and AI. Largely alone, he has built Cortex (the graph engine beneath Daxo), Daxo itself, Catalyst (a trading system), and a number of applications in between. This is not a pivot chased for a raise — systems and software are the work he keeps returning to.
Daxo began as something Daniel wanted — a dependable second brain, a digital right hand to rely on continuously, and with confidence, to hold the whole picture and answer it in plain language. It does that for Daniel now. This round is about letting other operators rely on it too.
Away from the screen, Daniel is developing a hobby farm in the Hunter region of New South Wales — a long-term, hands-on project built for the family and close to his heart. That is where the founder allocation is headed: not a splurge, but roots. It speaks to a founder settling in for the long run, in one place, rather than one angling for a quick exit.
Why the A$200K backs the founder
The platform is as complete as it is because Daniel has been able to give it his full attention. The founder allocation keeps that attention intact through launch — so the push from built to paying is not interrupted by a return to other work. It protects the project's single most productive input.
Track record
- Built and exited an accounting practice
- 15 years self-directed as a full-time trader
- 3 years full-time in AI — Cortex, Daxo and Catalyst shipped