FAQ

I focus on AI agent systems, pipeline automation, workflow orchestration, and MCP integrations. Typical projects include building conversational AI interfaces backed by real tools and data, automating multi-step business processes with AI, integrating AI capabilities into existing software stacks, and setting up data ingestion pipelines. If it involves making AI do useful work in a business context, it's likely a good fit.

It depends on scope. A focused single-agent build or pipeline integration can take one to two weeks. Larger projects involving multiple agents, complex workflows, and data migration typically run four to eight weeks. I always scope work clearly upfront so there are no surprises, and I break larger projects into milestones with working deliverables at each stage.

Both. I work with startups, small businesses, and established organisations. The common thread is that the client has a real business problem where AI can add measurable value. Company size matters less than having clarity on what you need and being ready to engage in the process.

I work on a project basis with clear scoping upfront. After an initial consultation, I provide a fixed-price proposal covering the agreed scope. For ongoing maintenance and support, I offer monthly retainer arrangements. I don't bill by the hour for development work because it creates the wrong incentives. You're paying for outcomes, not time.

Yes. A large part of what I do is integrating AI capabilities into systems you already use. Whether that's connecting to your CRM, ERP, internal databases, or SaaS tools, I build bridges rather than replacements. MCP (Model Context Protocol) is particularly useful here as it provides a standardised way for AI agents to interact with external systems.

Absolutely. AI systems aren't set-and-forget. Models evolve, APIs change, and your business needs shift. I offer maintenance retainers that cover performance monitoring, prompt refinement, integration upkeep, and incremental improvements. I also provide handover documentation if you prefer to manage systems in-house after deployment.

Seriously. I sign NDAs as standard practice, and I design systems with data privacy in mind from the start. This includes choosing appropriate model providers, understanding data residency requirements, and ensuring sensitive information is handled according to your compliance needs. You can read more about my approach on the confidentiality page.

I'm model-agnostic. I work with Anthropic (Claude), OpenAI, Google Gemini, and open-source models depending on what's best for each use case. For orchestration, I use tools like n8n, CrewAI, and custom frameworks. Deployment can be on SaaS platforms, self-hosted infrastructure, or cloud environments. The right answer depends on your requirements around cost, privacy, performance, and existing infrastructure.

I'm based in Jerusalem, Israel (IST / UTC+2). I work remotely with clients worldwide. Most of my communication happens asynchronously via email, project management tools, and shared documents, with scheduled video calls for key check-ins and reviews. The async-first approach works well across time zones and keeps projects moving efficiently.

The best starting point is to submit a project spec. It doesn't need to be detailed — even a rough description of what you're trying to achieve is enough to start the conversation. I'll review it and get back to you with initial thoughts, questions, and next steps. You can also get in touch directly if you'd prefer to have a conversation first.