From zero to one.
Governed at scale.
0n1.ai is a data, BI and agentic AI platform for Asia-Pacific. We start from the business outcome — BI and data strategy, AI advisory, engineering that reaches production — then, when you want it permanent, we establish a sovereign platform with governance, observability and audit built into the substrate, not bolted on.
Consulting that ends in an outcome, not a deck.
Four branches, one operating philosophy: name the business outcome first, prove it with the smallest credible piece of work, then industrialise it. Engage us for a single branch, or for the whole arc from strategy to production.
Data & BI Consulting
Strategy through implementation across the data estate. Tool-agnostic by conviction: Tableau, Power BI or Looker, on whichever architecture fits.
- Data & BI strategy, operating model & adoption
- Semantic layer & governed metric definitions
- Dashboard rationalisation & platform migration
AI Consulting
Where AI actually pays, and where it doesn't. Honest value cases, target architecture and a sequenced roadmap your CFO can read without a translator.
- AI opportunity assessment & use-case portfolio
- Value cases & benefits realisation
- Target-state architecture & build-vs-buy
AI Engineering
Build teams that ship to production, not to a demo. Agents, predictive ML, computer vision and NLP, with the evaluation discipline that keeps them working after go-live.
- Agentic systems, RAG & knowledge pipelines
- Predictive ML, computer vision & NLP
- MLOps / LLMOps & evaluation harnesses
AI Governance
The assurance layer that lets a regulated board approve AI — operating model, policy, evaluation and audit, designed against the frameworks you actually answer to.
- Governance operating model & decision rights
- Risk tiering, policy & impact assessments
- Regulator-ready evidence & audit trails
Data & BI is a strategy problem, not a licence problem.
Most BI programs don't fail on the tool. They fail on definitions nobody agrees with, ownership nobody holds, and numbers nobody trusts. We treat data & BI end-to-end, then implement it on whichever platform you're standing on.
B1 Decision mapping & BI strategy
Start from the decisions your business actually makes — who makes them, on what cadence, with what evidence — and work backwards to the reporting that serves them.
B2 Semantic layer & metric definitions
One governed definition of revenue, churn, margin and utilisation — modelled centrally so every dashboard, notebook and agent answers the same question the same way.
B3 Platform & pipeline implementation
The unglamorous layer beneath the dashboard: dimensional modelling, ELT pipelines, warehouse or lakehouse design, freshness SLAs and cost control.
B4 Migration & rationalisation
Consolidating sprawl — hundreds of legacy reports triaged, retired or rebuilt across Tableau, Power BI, Looker or on-prem to cloud — without losing the numbers people trust.
B5 Adoption, enablement & literacy
Self-service guardrails, certified datasets, training and a support model — so the platform survives the day the consultants leave.
B6 AI-ready foundations
The same governed model that powers your dashboards is what copilots and agents will query. Done properly, BI is the cheapest AI enablement you will ever do.
Consulting that leaves a platform behind.
Most engagements end with a deck and a handover, and the point solution decays from there. Ours can end with a durable operating layer instead: one governed data plane, one agent runtime, one control plane — so every use case you launch next inherits the same foundations. Optional, and only when it earns its keep.
Sovereign Data Plane
A lakehouse-pattern foundation deployed in your cloud, your region, or on sovereign infrastructure — your data never becomes someone else's asset.
- Lakehouse & streaming architecture
- Unified catalog, lineage & access control
- Cloud migration & modernisation (GCP, AWS, Azure)
- Onshore & air-gapped deployment options
Agent Runtime
Model-agnostic by design. Run frontier, open-weights or fine-tuned models — and swap them without re-platforming — behind one orchestration layer.
- Multi-agent orchestration & tool use
- RAG & enterprise knowledge pipelines
- Computer vision, NLP & predictive ML
- Evaluation harnesses & regression gates
Governance Control Plane
Policy, risk and assurance enforced at runtime. Every agent action is permissioned, traced and attributable — from prototype to production.
- Risk tiering & human-in-the-loop controls
- Full-fidelity agent tracing & audit trails
- Guardrails, kill-switches & rollback
- Regulator-ready evidence packs
Sovereign GPU capacity, run like a utility.
Agentic AI is bottlenecked by compute, not ideas. 0n1 designs and operates the accelerated-compute layer beneath the platform — from sovereign GPU clusters and AI factories to research-grade HPC — so scarce accelerators are shared, scheduled and governed, not hoarded.
C1 Sovereign AI factories
On-shore GPU clusters for government, defence and regulated industries — designed for data residency, air-gapped operation and open-weights model serving at scale.
C2 Research & scientific HPC
Hybrid HPC + AI environments: batch simulation and training workloads on shared accelerators, with fair-share scheduling across departments/business units and projects.
C3 GPU orchestration & fractional scheduling
Kubernetes-native GPU pooling, quotas, pre-emption and fractional allocation so inference, fine-tuning and training jobs coexist — driving utilisation up and cost per token down.
C4 Compute governance & FinOps
Per-team, per-agent GPU accounting; chargeback and showback; capacity forecasting; and policy that ties compute entitlement to the risk tier of the workload it runs.
Governance is the product — not the paperwork.
Tooling, observability and deployment controls are table stakes. 0n1 builds the full governance operating model around them: accountability, regulatory alignment, assurance and culture — so your board can say yes to agentic AI with evidence, not hope.
G1 Operating model & accountability
AI governance councils, decision rights, risk appetite statements and role-level accountability mapped to your existing three-lines-of-defence model.
G2 Policy & risk tiering
Use-case classification, impact assessments and human-oversight thresholds — codified as policy-as-code so they're enforced, not just filed.
G3 Evaluation & assurance
Golden datasets and offline eval suites per use case; LLM-as-judge scoring calibrated against human raters; agent trajectory and tool-use evals; adversarial red-teaming; bias and robustness testing — all wired into CI as regression gates, then re-run continuously in production.
G4 AI observability & audit
Every prompt, retrieval, tool call and decision traced end-to-end with OpenTelemetry-native instrumentation. Quality signals (groundedness, hallucination rate, refusal rate), performance (latency, token cost, GPU utilisation) and drift are monitored together — queryable by risk teams, exportable for regulators.
G5 Deployment & lifecycle controls
Staged rollout, canarying, drift monitoring, incident response runbooks and decommissioning — the full lifecycle, governed.
G6 Data sovereignty & privacy
Residency guarantees, open-weights deployment on sovereign infrastructure, and privacy-by-design pipelines for regulated data.
Any agent. Your platform.
Because governance and data live in the platform, agentic use cases become configuration — not twelve-month rebuilds. Proven patterns from a track record across telco, media, advertising, energy and retail.
Network & data modernisation
Enterprise data migration to cloud for a major telco — absorbing a 2× workload increase while cutting total cost of ownership by 30%+.
Real-time intelligence at scale
ML-powered video intelligence serving 1M+ daily active users across tier-1 mastheads, plus Vertex AI MLOps delivering material cost savings for a digital-advertising leader.
Computer vision in production
99% vehicle-detection accuracy (a 20%+ uplift) recovering revenue leakage for a tier-1 parking operator; automated solar roof-detection cutting manual effort by 17%.
Built in APAC, for APAC.
- Sovereignty firstOpen-weights & onshore hosting options for government, defence and regulated industries
- Regulatory fluencyAPRA, IRAP, MAS and IMDA frameworks treated as design inputs, not afterthoughts
- Partner ecosystemDeep alliance experience across Google Cloud, AWS and Databricks ecosystems
- Regional deliveryExecutive advisory in-market across ANZ and ASEAN, delivery that follows the sun
Outcomes, measured.
Founded on 15+ years leading data, AI and cloud engagements across APAC — from CTO of a machine-learning product company to scaling one of ANZ's leading cloud & data practices.
Detection accuracy
Computer-vision uplift of 20%+ that closed revenue leakage for a tier-1 parking operator.
TCO efficiency
Telco data-platform migration to cloud, absorbing a 2× workload increase.
Daily active users
ML-powered video platform in production across major ANZ media groups.
Practice scaled
Billable consulting team grown and coached across data, AI/ML, CV and NLP.
Programs owned
End-to-end ownership of delivery, governance, risk and budget at executive level.
Manual effort removed
Automated solar roof-detection with segmentation models on AWS SageMaker.
Establish your platform. Then run any agent on it.
A 90-minute executive briefing: your data estate, your regulatory posture, and a candid view of what agentic AI can — and can't — do for you in the next two quarters.
Request a briefing — future@0n1.ai