Thirteen years offshore taught me what happens to clever plans when they meet a fabrication yard. I bring that discipline to AI: seminars that train your team on real cases, advisory days that map your operation, and systems delivered with partners I trust. No hype. Nothing I cannot defend in a meeting.
Approved for HRDA funding for eligible Cyprus employers. Built for operators, not spectators: every session runs on your industry's real cases. Delivered directly and through certified training partners.
Where blockchain genuinely earns its place in your operation, and where it does not. Real applications, practical tools, zero evangelism.
Ask about this programme →AI across the energy value chain: forecasting, operations, documentation and reporting. Tools your engineers can run the next morning.
Ask about this programme →AI applied to chartering, port calls, compliance and the back office. Taught by someone who has run marine logistics for a living.
Ask about this programme →Design and deploy agents that take real work off your team: tracking, documentation, exceptions. Automation that holds under load.
Ask about this programme →Ask for the current schedule or an in-house date. Programmes run in English or Greek.
You bring the people who know the work. I bring the method. In one session day we walk your processes, find where AI actually pays, and write it down properly.
I take a limited number of advisory days each quarter. Training-linked work can run through HRDA funding. Everything else is priced on scope, agreed before we book.
Request an advisory day →Deliverables in writing within three working days of the session.
The path is deliberate. Teams that learn first make better automation decisions, and better clients.
Your team learns on your cases, in a funded programme where eligible. AI stops being a rumour.
An advisory day turns interest into a scoped plan with priorities and acceptance criteria.
Vetted delivery partners build under my architecture. You get the right builder, not the nearest one.
Systems go live with verification, handover and a clear owner. I stay on the strategy.
I get paid for judgement, teaching and architecture, not for selling you developer hours. That keeps my advice clean.
A sample from the register. Sector labels, not client names: references belong in meetings, not on websites. Statuses are real. IN SERVICE means running today.
| Tag | System | Sector | Status |
|---|---|---|---|
| TG-SYS-01 | Invoice Processing Classifier Supplier invoices classified and routed automatically, F1 0.94 | Shipping and services | Pilot |
| TG-SYS-02 | Lookbook Product Extractor Catalogue pages become structured product data for ordering | Retail distribution | Pilot |
| TG-SYS-03 | Messaging Invoice Capture An invoice photo in a chat becomes a ledger row | SME finance | Demo |
| TG-SYS-04 | M365 Workflow Automation Eight live workflows across mail, tasks and reporting | Consultancy operations | In service |
| TG-SYS-05 | Research Paper Evaluator HSE papers screened and scored against custom criteria | HSE and R&D | Demo |
| TG-SYS-06 | Email Triage and Drafting Local classifier routes mail, AI drafts the replies worth drafting | Operations | In design |
Builds are delivered with vetted partners under my architecture. Ask for the full register in a meeting.
Founder, Toolguide AI · Industrial Project Manager · HRDA-Registered Trainer
I came up through offshore oil and gas: Aberdeen, Saudi Arabia, the East Med. Subsea fabrication, load-outs, marine logistics. Projects where a missed tolerance costs six figures and a missed deadline costs a vessel day. That world does not forgive guesswork.
AI gets the same treatment. I test everything on my own operation first: my invoicing, my email and my reporting run on systems I built. What I teach in a seminar ran in production the week before.
Toolguide AI is the teaching and advisory practice of Synnos A Energy Solutions Ltd. Based in Nicosia, working across Cyprus and the East Med, in English and Greek.
Short, practical writing on AI in real operations. No newsletter machinery, no spam, ever.
Every failed pilot I have seen died the same way: it was built to impress in a meeting, not to survive a Tuesday. The demo answers a clean question with clean data. Your operation has neither. The fix is boring and it works: pick one process, define what done means before you build, and put acceptance criteria in writing. If a vendor resists acceptance criteria, you have learned what you needed to know.
Reconciliation, report compilation, data entry between systems. The repeat offenders across Cyprus SMEs, and what each one costs you.
Thirteen years around vessels and ports, every hype cycle included. What AI actually changes in shipping and marine logistics.
Full notes publish here first.
Tell me what you are trying to fix. I reply within one working day with an honest read: what AI can do for it, what it should cost, and whether I am the right person. If I am not, I will say so and point you somewhere better.
I reply within one working day. Talk soon.