1990s · Citrix
Work moves beyond the office
Virtualisation changed where work could happen. Adoption depended on partners who could explain and implement the shift.
Carey Bowden · EMEA Ecosystem Leader
Experience. Integrity. Judgement. Proven delivery.
Thirty years building the partnerships, operating models and governance that turn a platform into something enterprises actually use. Now I build agentic AI systems myself.
30 yearstechnology adoption
189 partnersacross EMEA
1 working agentbuilt, not presented
ORCHESTRATORJeevesLIVE SYSTEM
Memorycontext + tasks
Researchnews + files
Markets5 paper books
Buildspec → code
Careerradar + pipeline
CommsWhatsApp + email
Carey approval required
Technology creates opportunity. Ecosystems make it happen.
Five technology waves
The technology changes. The human problem doesn't.
My career has been spent where technical possibility meets commercial reality: making the human case for change, building the route to market, and scaling it across countries and cultures.
2000s · Symantec
Trust becomes infrastructure
Enterprise security moved from product conversation to board-level dependency, demanding scale, clarity and confidence.
2010s · NVIDIA
Compute changes what is possible
Accelerated computing opened new enterprise markets. The channel had to learn a new language and a new value story.
2020s · AWS
Skills become the adoption layer
Across 50+ countries and 189 partners, training turns cloud intent into deployed customer capability.
Now · Agentic AI
Software begins to act
The next adoption challenge is not simply capability. It is autonomy, operating design and accountable human control.
The proof point
Jeeves is a working agentic system, not a deck about one.
I started with no coding experience. Jeeves now remembers, researches, monitors, drafts, builds and asks permission before high-stakes action. It operates across seven services with an explicit human approval layer.
7connected services
5paper traders live
16+routed intents
Read the world
- Fresh news & research
- Momentum & market data
- Insider and catalyst signals
- Memory & prior decisions
→
Separate the voices
- Analyst makes the case
- Sceptic challenges it blind
- Risk Officer rules on both
- Tests try to break separation
→
Human approval
1Propose the action2Show evidence or diff3Wait for Carey4Execute and audit
Architecture simplified from Jeeves Journey Map v15 · September 2026
SHIP'S LOG / SELECTED SYSTEMSSTATUS: OPERATING
Reads the world
Daily briefing, inbox triage, research capture and a memory layer that prevents repetition.
Watches markets
Five specialist paper-trading books with separated Analyst, Sceptic and Risk Officer roles.
Builds himself
Draft, confirm, ship or abort through WhatsApp. No change reaches production without approval.
The destination
Applied AI leadership: practitioner credibility joined to ecosystem scale and board-level governance.
Enterprise scale
From strategy to customer behaviour.
For a 24,000-customer EMEA population, I designed a three-layer Generative AI training motion: relationship-led engagement, AI-enabled scale and partner conversion—governed by real utilisation and paid-conversion gates.
24kcustomers in the addressable base
14.5kreceived a GenAI training offer
65%account-manager engagement
811registrations generated
532attendees across EMEA
34events delivered
189training partners
50+countries in the ecosystem
Earlier attempts had stalled on data quality and internal buy-in. The answer was not a larger campaign; it was an operating model that connected account-team trust to partner delivery capacity with explicit rules.
The results supported a 4,255-place next-phase capacity plan. Relationship-led and scale-led motions work together when governance is disciplined and reporting is honest.
The control boundary
Autonomy should expand. Accountability should never disappear.
Most organisations already have AI in production. Far fewer have redesigned authority around systems that can decide and act.
Building Jeeves exposed the quiet failures a policy document misses: checks that report green without biting, role separation that exists only on a diagram, and autonomous behaviour whose confidence is not evidence.
AI AUTONOMYHUMAN ACCOUNTABILITY
Gather evidenceGenerate optionsChallenge assumptionsDraft the action
CONTROL BOUNDARY
Set the mandateReview the reasoningAccept the riskAuthorise execution
Draft it. Show the diff. Wait for explicit approval. Ship.
Advisory · NED · AI leadership
Practitioner credibility, ecosystem scale and governance instinct.
I am building toward a senior AI leadership chapter and board roles with organisations serious about putting AI to work responsibly.
Ecosystem strategy
Partner networks that produce adoption and commercial outcomes, not signed agreements.
Applied AI
Hands-on experience building, testing and operating a real agentic system.
Enterprise GTM
Three decades navigating hyperscaler, vendor, partner and customer incentives.
Board governance
Clear mandates, separation of roles, evidence gates and explicit accountability.
Start a conversation