What I build

AI first, and everything it needs to be useful: the product, the integrations and the cloud it runs on.

Each one is described below in full. On a project, most of them combine: an agent needs a product around it, a product needs a cloud under it.

Agents that plan tasks, use tools, call your APIs, update your CRM and hand off to a person when they should. Built with LangGraph, OpenAI and Claude, with clear permissions, failure handling and monitoring so they keep working after launch.

Assistants that answer from your documents, knowledge base or database and show their sources. Chunking, embeddings, vector search, reranking and evaluation sets that measure answer quality instead of guessing.

Support, sales and internal chatbots on your website, in Slack or inside your product. Grounded in your data, with guardrails, escalation to humans and conversation analytics.

Adding AI to the tools you already use: prompt design, structured outputs, function calling, model selection and cost control. Automations in n8n and Make that connect your CRM, email, forms and databases, with human approval where it matters.

Custom web applications built with React, Next.js and Node.js: dashboards, user roles, authentication, payments, admin panels and integrations, designed to scale.

Custom CRMs and integrations with existing platforms: sales pipelines, lead scoring, automated follow ups and reporting. AI can be added on top for lead qualification and summaries.

Fast, well structured marketing sites and online stores. Custom coded sites, WordPress (including Elementor and WooCommerce) and Shopify stores, built for speed, SEO and easy editing by your team.

Product and website design from wireframes to polished interfaces in Figma: user flows, design systems and responsive layouts that developers can build exactly as designed.

iOS and Android apps with React Native, connected to the same backend and AI features as the web product.

Deployment on AWS and Azure, also Vercel and similar platforms where they fit. Docker, CI/CD pipelines, monitoring, logging, backups and cost control, so what we launch stays fast, secure and affordable to run.

Short answers to what clients ask before a first call.

No. AI is the focus, but we also design and build websites, web apps, CRMs, WordPress and Shopify sites, and mobile apps, and we deploy and maintain them on the cloud.

Every AI system is tested against real questions before launch, with evaluation sets, guardrails and logging. After launch we monitor accuracy, speed and cost, so problems are caught early.

Your data stays in your accounts and is only used to serve your application. We can deploy on your own AWS or Azure account if you prefer full control.

Yes. All code, designs and documentation are yours at handover.

With me directly. For larger builds a dedicated team supports design, development and deployment, and you still have one point of contact.

Yes. Most projects connect to systems you already use, through their APIs or automation platforms like n8n and Make.

With a short call. You explain the workflow or product, and I reply with an approach, a realistic timeline and a price range.