Connect your data, describe your goal, get a deployed model. An autonomous AI data scientist that profiles data, writes its own training code, validates results, and ships production endpoints — end-to-end, in hours not months.
Decisions that depend on data — churn, forecasting, demand, defects, revenue — shouldn't take months. Today they do, because shipping a real model still requires notebooks, MLOps, infrastructure, and a scarce expert team that can frame the problem, iterate, and debug.
Legacy AutoML (DataRobot, H2O) spits out model leaderboards. It still needs an expert to frame the task, engineer features, debug failures, and operationalise the winner. That's where months go.
Even a simple model needs Jupyter, packages, CI, a registry, monitoring, a deployment surface, and a human stitching them together. Most teams can't carry that stack just to answer one business question.
Data scientists want to ship more models, faster. Business users want to ask a question in plain English and get a defensible answer. Nothing in the market serves both without compromise.
OctOpus runs the same loop a senior data scientist runs, but in hours. It writes its own per-experiment training code, debugs failed runs, compares approaches, and delivers a production endpoint plus the report a stakeholder can act on.
Drop a spreadsheet, database, or file. State the business question — churn, revenue, demand, defects — in plain English.
OctOpus profiles the data, frames the task, picks features, chooses a model family, and prepares a deployment path.
Runs experiments, repairs failures, compares approaches across model families until results are production-ready.
Ships a deployed model, a defensible report, and actionable outputs for whoever needs to make the call.
All revenue inbound. No outbound sales. Paid pilots converting to longer-term contracts. Numbers and named accounts shared privately under NDA in the live deck.
Active discovery and paid pilots across the sectors below. Named accounts shared privately on a call.
All Fortune-500-class buyers. All inbound — they reached out to us.
OctOpus automated the bulk of a manufacturing partner's QA reporting pipeline — eliminating manual analyst work while maintaining accuracy. Hours of model work, not months. A single operator now ships what a team used to scope. Full case study and metrics available under NDA.
Coding agents write functions. OctOpus owns an entire data-science loop — framing, iteration, repair, validation, deployment — backed by domain priors. That's the difference between "wrote some Python" and "shipped a decision."
Clean SaaS ladder today. Long-term, as reliability compounds, pricing shifts from seats to outcomes — payment for a delivered forecast or a deployed model.
A decade in applied AI, optimisation, and automation — operating where research-grade ML meets production decisions at global scale. OctOpus isn't a side project; it's the productisation of work that's already saved a Fortune-500 carrier real money every day.
10+ years across applied AI, optimisation, and automation. PhD-level ML rigor paired with production deployments at one of the world's largest logistics carriers. Currently founding two companies in parallel — OctOpus and OneNine — while leading operations-research work at Maersk.
Already live, already monetising, already pulling Fortune-500-class inbound. The pre-seed removes the day-job dependency, brings in the first 1–2 hires, and turns the enterprise pipeline into category leadership before the window closes. Round size and terms shared privately.
We share the live deck with investors directly. A line or two about you and your firm so we can send the right context back.
We've received your request and pinged the founder. You'll hear back directly. In the meantime, the deck is open below.
Open the deck →
OctOpus
