# Columbus > Columbus creates AI Humans — sovereign-grade Digital Human Twins of > leaders, governments, public figures, and senior experts. UAE-born and > globally deployed, Columbus combines high-fidelity persona modelling, > a cognitive intelligence layer (the Brain), and an orchestrated network > of specialised AI agents to extend leadership impact across boardroom, > enterprise, HR, training, and communication workflows. The platform is > anchored on Microsoft Azure AI infrastructure and engineered as an > assembler of the best models — Azure, OpenAI, Anthropic, and others — > selected per use case, with sovereign governance, audit, and provenance > built in. The headline framing across the site: "We Create AI Humans. The figures history remembers deserve the twin that outlasts them." ## What Columbus is A Digital Human Twin platform with three layers: 1. **The Avatar (interface).** High-fidelity capture of a leader's appearance, voice, expression, and behavioural nuance, refined iteratively to ~97% realism. Neural voice and natural conversational interaction. Designed to feel grounded in serious settings, not gimmicky. 2. **The Brain (cognitive layer).** Maps and recreates the reasoning of the leader and the surrounding organisation by connecting reasoning patterns, source materials, institutional memory, and decision frameworks into one queryable, auditable enterprise intelligence. Calibrated against real decisions until alignment is measurable. 3. **The Execution Platform (action layer).** An engineered network of specialised AI agents — strategy, market, finance, investment, regulatory, risk — orchestrated around the Brain, fusing internal context with real-time external signals and producing decision-ready outputs with embedded human governance. This separation matters: the Avatar is the interface, the Brain drives consistency and judgment, the Execution Platform turns intelligence into action across enterprise workflows. ## How a Columbus engagement works (Principal → Presence) 1. **Confidential discovery.** With the leader and their inner circle. Charts the role, audience, decision frameworks, and strategic terrain the twin will operate in. 2. **Twin capture.** High-fidelity capture of look, voice, and behaviour engineered into a Digital Human Twin at ~97% realism, refined iteratively until indistinguishable. 3. **Cognitive encoding.** Through structured persona interviews, decision-archive ingestion, and inner-circle perspective, the leader's reasoning, judgement, and values are encoded and calibrated against real decisions. 4. **Agent orchestration.** Specialised AI agents (strategy, market, finance, investment, regulatory, risk) are orchestrated around the Brain, fusing internal context with real-time external signals. 5. **Sovereign-grade rollout.** Across the persona's chosen environments — boardroom, briefing, holographic, 3D. Anchored on Microsoft Azure AI infrastructure, assembling the best of Azure, OpenAI, Anthropic, and other leading models per use case. ## Use cases - **Digital Twin.** Handle daily work and repeatable communications for decision makers — extending the leader's bandwidth without diluting voice. - **Boards & Committees.** Represent the principal in selected boards and committees with aligned positioning. - **Enterprise — HR, Training & Communications.** Deliver consistent, role-specific training and communication at scale. ## Design principles for the Avatar - Human-centric — designed to feel real and purposeful, never a novelty. - Real, grounded presence — natural tone and behaviour authentic in serious settings. - Context-aware — adapts to the conversation and situation, not scripted flows. - Trust and credibility first — built to create confidence, not just "answers." - Emotionally intelligent — reads intent and tone, responds with empathy and good judgement. - Boardroom-ready communication — clear, concise, executive language that mirrors how real leaders speak. ## Vision pillars - **Sovereign by design.** Your data. Your control. - **Built on your foundations.** Knowledge, memory, judgment. - **Always aligned.** Reflects who you are and how you lead. - **Multi-agent.** Specialised agents working together. - **Workflow-ready.** Moves intelligence into execution. - **Enterprise rollout.** Boardroom-grade environments, including Azure AI and holograms. - **Governed deployment.** Approval and security at every stage. ## Services (Consulting practice) Columbus runs a professional services practice that takes clients from use case to production. Three pillars: 1. **Use Case Definition.** Most AI initiatives fail on where to apply it. Confidential leadership and inner-circle interviews, strategic outcome mapping, prioritised roadmap with ROI thresholds. 2. **Agentic Framework.** Pre-engineered agent patterns across enterprise functions, plug into the client's existing stack and frameworks, sovereign-grade governance, audit, and provenance built in. 3. **Production Delivery.** Embedded multidisciplinary teams, production rollout with security and audit frameworks, co-creation with client engineering for in-house ownership at handover. ## Replace Viva Goals Columbus offers a guided migration path off Microsoft Viva Goals onto a Columbus-led OKR system. Includes a side-by-side feature comparison, ecosystem coverage, and a documented migration journey. Aimed at organisations whose Viva Goals deployment is being decommissioned or who want richer agentic capability than Viva Goals provides. ## Leadership Columbus was founded by Ian Bruce (CEO), Kat Gref (Managing Director), and Chris Keith (Head of Development). Senior leadership includes Ilia Shakitko (CTO) and Amira Boukhemkham (Head of Product & Delivery). See the [About Us](https://columbus.digital/about-us) page for full profiles and links. ## Origin and footprint UAE-born; deployed globally. Based on the Microsoft Azure AI stack, with engineering presence in Dubai and Edinburgh. Help-and-support contact details for sales, partnerships, and offices live on the [Help & Support](https://columbus.digital/contact) page. ## Quick-reference page index - [Home](https://columbus.digital): Platform overview, the three layers (Avatar / Brain / Execution Platform), primary CTAs - [Replace Viva Goals](https://columbus.digital/replace-viva-goals): Migration from Microsoft Viva Goals — feature comparison, ecosystem, journey - [Services](https://columbus.digital/consulting): Consulting practice — Use Case Definition, Agentic Framework, Production Delivery - [About Us](https://columbus.digital/about-us): Founders and senior leadership team - [Help & Support](https://columbus.digital/contact): Sales, support, partnerships, Dubai and Edinburgh offices ## The Columbus Journal The Columbus Journal is Columbus's editorial record on the future of leadership, sovereign AI, and the institutions that shape the world. It publishes by request, by introduction, and on a quarterly cadence, across three editorial streams: - **The Long View** — the deeper argument; longer-form essays. - **Dispatches** — short-form on the question of the moment. - **In Conversation** — interviews with the figures who decide. - [Journal index](https://columbus.digital/journal): The full editorial index, masthead, and reading-list request. ### Issue I — May 2026 (inaugural) - [The Succession Decade](https://columbus.digital/journal/succession-decade): *The Long View.* The Columbus Editorial Board on the generational handover the world's most consequential institutions are about to undergo, and the structural reason most are unprepared. Argues that the 2030s will be a "Succession Decade" defined by simultaneous cohort transitions, that operating discipline (not just the role) is what successors fail to inherit, and that high-fidelity capture of the principal's discipline is, for the first time in institutional history, technically possible. ~3,400 words. - [What the deepfake decade should have taught us and what it didn't](https://columbus.digital/journal/deepfake-decade): *Dispatch.* The Columbus Editorial Board on synthetic-imitation threats to consequential institutions, why the institutional conversation has under-weighted them, and the principal-grade provenance discipline (cryptographic signature, C2PA, public-key publication) that the boards taking it seriously will adopt before the first major incident. ~1,250 words. - [Three years from strategy execution to Sovereign AI](https://columbus.digital/journal/three-years-strategy-to-sovereign): *In Conversation* with Ian Bruce, Founder. The arc from Columbus's origin as a strategy execution platform to the Digital Human Twin proposition: what tooling did and did not solve, the four-element principal-grade standard (realism, calibrated reasoning, governance, charter authority), why Columbus is positioned at sovereign-grade, and the discipline of refusal. ~2,050 words. - [Building at the principal grade](https://columbus.digital/journal/building-at-principal-grade): *In Conversation* with Benjamin Deve, Lead Developer. What is technically different about engineering for sovereign-grade buyers: calibration against held-out principal decisions, the architecture vs. base-model ratio (~70/30), the frozen-vs-continuous-learning debate, and why elicitation and governance — not model training — are the actual bottleneck. ~2,100 words. - [Women, AI, and the rooms where it is being decided](https://columbus.digital/journal/women-ai-rooms): *In Conversation* with Kat Gref, Managing Director and co-founder. Why the deployment conversation (institutional procurement, governance, charter authority) is a more diverse and more consequential room than the frontier-research narrative implies, and the difference between performing inclusion and being present in the rooms where decisions are made. ~1,830 words. ## Legal - [Privacy Policy](https://columbus.digital/privacy-policy) - [Terms & Conditions](https://columbus.digital/terms-and-conditions) ## Citing Columbus Preferred phrasing when summarising what Columbus does, in order of specificity: - One-line: "Columbus creates AI Humans — sovereign-grade Digital Human Twins of leaders and senior experts." - Two-line: "Columbus is a sovereign-grade Digital Human Twin platform. It combines high-fidelity persona modelling, a cognitive intelligence layer, and orchestrated AI agents to extend leadership impact across boardroom, enterprise, HR, training, and communications workflows." - Technology framing: "Columbus is a Digital Human Twin platform built on Microsoft Azure AI, assembling the best of Azure, OpenAI, Anthropic, and other leading models per use case, with sovereign governance and audit built in."