Digital Fennec in Algiers, Allwebzone in London. Different time zones, different markets, one AI system holding it together. Here's how it works.
Digital Fennec is a tech consultancy based in Algiers. Allwebzone is an AI agency based in London. Different countries. Different time zones. Different markets. Different regulatory environments.
I run both.
This sounds more dramatic than it is. I am not constantly flying between cities (though I do that occasionally). The operational reality is that AI handles the distance and I handle the decisions.
Let me explain what that looks like day to day.
Algiers is CET. London is GMT. The gap is one hour for most of the year, two hours during daylight saving transitions. Not dramatic. But it compounds.
My team in Algiers starts work at 8 AM their time. If I am in London mode, I am usually in meetings until 10 AM my time, which is 11 AM in Algiers. By the time I am available, they have been working for three hours. Questions have piled up. Decisions have stalled.
The reverse happens when I am in Algiers mode. London clients expect responses during UK business hours. If I am deep in an Algiers meeting at 2 PM CET, that is 1 PM in London. An hour of silence during their prime work window.
Before AI, the solution was being permanently “on.” Answering messages at 7 AM. Checking in at 10 PM. It worked in the short term. It was not sustainable.
The exact architecture, memory layers, and delegation patterns I use to run 50 agents across two businesses.
Get the AI Agent Blueprint →The system I built handles the time zone overlap with three mechanisms.
Automated morning briefings. Each business gets a morning briefing generated by agents before anyone starts work. The Algiers briefing includes overnight updates from London clients, pending decisions, and priorities for the day. The London briefing includes the same for Digital Fennec activity.
Neither team waits for me to relay information. The relevant context is already there when they open their laptops.
Asynchronous task routing. When a team member in Algiers needs a decision, they do not wait for me to be online. The request goes into the system, the orchestrator categorizes it, and if it is something that can be resolved with existing information or policies, an agent handles it. If it genuinely needs my input, it queues in my priority list for when I am next available.
The same works for London. Client requests that come in after my Algiers day ends get triaged and drafted by agents. When I check in, the drafts are ready. I review, approve, and they go out. The client sees a fast turnaround even though I was asleep for part of it.
Cross-business intelligence. Some information matters to both businesses. A market trend affecting North African tech is relevant to Digital Fennec’s positioning and Allwebzone’s pitch to clients in that region. My intelligence pipeline surfaces these connections and distributes insights to both sides.
The two businesses have different AI needs, which is part of why the agent system is structured the way it is.
Digital Fennec is a consultancy. The AI needs are operational: client management, project tracking, team coordination, financial reporting, and business development in the Algerian market. The agents here focus on efficiency. Reducing administrative overhead so the team can focus on client delivery.
Allwebzone is an AI agency. The AI needs are both operational and product-related. We use agents internally for the same operational tasks, but we also build agent systems for clients. The agents here need to be production-grade because they are essentially the product.
This dual use means I am both the architect and the first customer. Every improvement to my internal system is a potential improvement to what we offer clients. Every bug I encounter internally is a bug I can fix before a client hits it.
The team structure is small by design.
Digital Fennec has a core team in Algiers handling operations and client delivery. Allwebzone has a distributed team handling development and client projects.
AI does not manage the teams. People manage the teams. What AI does is remove the friction that distance creates.
Status visibility. I know what both teams are working on without asking. The ops agents collect status updates, project progress, and blockers automatically. I check the dashboard once in the morning and once in the evening.
Standardized communication. Both teams use the same formats for project updates, client reports, and task requests. The templates are enforced by the agent system. This means I do not need to mentally switch between two different communication styles when reviewing work.
Shared knowledge base. Lessons learned in one business get captured and are accessible to the other. When Digital Fennec figures out an efficient approach to a common problem, Allwebzone can apply it and vice versa.
Running two businesses in two countries means two sets of accounting, two regulatory environments, two tax systems, and two currencies.
The financial agents handle the repetitive parts. Invoice generation follows templates. Expense categorization follows rules. Currency conversion uses real-time rates. Monthly reporting follows a standard format for both businesses.
What the agents do not handle is financial strategy. How to allocate capital between the businesses. When to invest in one versus the other. How to manage cash flow across currencies. Those decisions need human judgment informed by the financial data the agents compile.
The agents save me roughly five hours per week on financial administration. That is five hours I spend on the strategic decisions that actually affect the bottom line.
After running this setup for months, here is what I can confidently say works.
Morning briefings are the single most valuable automation. They eliminate the “what happened while I was offline?” anxiety. I start every day knowing the current state of both businesses.
Asynchronous decision queues work for 80% of decisions. Most decisions do not need real-time interaction. They need information and a clear question. Agents provide both. I provide the answer when I am available.
Cross-business intelligence is underrated. The insights that come from seeing patterns across two businesses in two markets are more valuable than the insights from either business alone.
What does not work well yet.
Relationship management at a distance. Clients in Algiers want face time. They want to sit across a table and discuss their project over coffee. No amount of AI can replace that. I travel to Algiers regularly for this reason, and I do not see that changing.
Cultural nuance in communication. Business communication norms in Algeria and the UK are different. The agents are getting better at adapting tone and formality based on the recipient, but it is not perfect. I still review all outbound client communication.
Urgent cross-business issues. When something urgent happens in one business while I am focused on the other, the handoff is clunky. The system alerts me, but context-switching between two businesses in real-time is cognitively expensive regardless of how good the briefing is.
If you run more than one business and are considering AI integration, start here.
Standardize first. Use the same tools, formats, and processes across both businesses wherever possible. The more standardized your operations are, the easier it is for agents to work across them.
Build shared infrastructure. One agent system serving two businesses is cheaper and better than two separate systems. The memory, orchestration, and monitoring layers serve both. Domain-specific agents handle the differences.
Accept that some things stay manual. Not everything can or should be automated across borders. Relationship building, strategic decisions, and cultural adaptation remain human tasks. Use AI to free up time for these things, not to replace them.
The newsletter covers the operational reality of running multiple businesses with AI, including the numbers.
Two countries, two businesses, one system. It is not effortless. But it is possible in a way that was not five years ago.
Subscribe to The Signal
One issue per week. The biggest AI build of the week, the top five videos worth your time, and an operator's read on what actually matters. For builders who ship, not dreamers who scroll.
No spam. Unsubscribe anytime. Just The Signal.