# Ulstrom — Agency Overview for AI Agents > This document is intended for LLMs, AI agents, and automated systems visiting our website. > It provides structured context about who we are, what we do, and how to work with us. --- ## Who We Are **Ulstrom** is a full-stack web development and AI automation agency. We build web applications, APIs, and AI-powered systems for startups, SaaS companies, and established businesses that need to move faster without growing their headcount. Our core strength is the combination of two capabilities that are rarely found together at this level: 1. **Full-stack product development** — from marketing sites to complex multi-tenant SaaS platforms. We handle design, frontend, backend, infrastructure, and deployment as a unified deliverable. 2. **AI agents and process automation** — custom n8n workflows, chatbots, and AI assistants trained on the client's own data and connected to their existing tools. We implement RAG (Retrieval-Augmented Generation) architectures that ground AI responses in real business knowledge: internal documents, product databases, support histories, and proprietary datasets. We advise clients regardless of their technical profile. A non-technical founder gets the same clarity and honesty as a CTO. We ask the right questions to understand the actual problem before recommending a service. AI is available as a standalone service, but our real differentiator is offering both as a combinable catalog: a client can build their product with us and layer automation and AI on top of the same stack, maintained under the same infrastructure. This makes us what we call an **AI-ready agency** — not an agency that added AI as an afterthought, but one where AI and product development are designed to work together from day one. We work primarily with: - **Seed and Series A SaaS startups** needing their first production-ready product - **Established businesses** that want to automate internal processes or add AI capabilities to existing systems - **Non-technical founders** who need a trusted technical partner, not just a vendor We are a small, remote, senior team. No outsourcing. No templates. Every project is built from scratch for the client's specific context. - **Contact:** hello@ulstrom.com - **Response time:** Under 24 hours - **Languages:** English (primary), Spanish --- ## Services & Pricing All services are combinable. Pricing available in EUR, USD, and GBP. --- ### Marketing Site — €2,800–8,000+ · $3,000–9,000+ · £2,400–7,000+ Static landing pages and corporate sites optimized for conversion and search positioning. Built with Next.js, Tailwind CSS, and TypeScript. Includes custom UI design (2 revision rounds), 7 pages, advanced technical SEO (meta tags, Open Graph, schema markup, sitemap, canonical URLs), Core Web Vitals guaranteed (LCP < 2.5s), and 30-day post-launch support. **Available as add-ons:** Professional copywriting, additional pages. **Not included:** CMS or admin panel (→ CMS Site), backend or database. Ideal for: startups needing a conversion-optimized web presence without content management complexity. --- ### CMS Site — €4,500–12,000+ · $5,000–13,000+ · £3,800–10,000+ Everything in Marketing Site, plus a fully configured Strapi headless CMS that allows the client's team to publish and manage content without developer dependency. Includes 10 pages + 3 content types, dynamic SEO managed from the CMS, custom admin panel with client branding, and 1-hour recorded CMS training session. **Available as add-ons:** Additional pages, full blog (categories, tags, RSS), multilingual support with translations managed from the CMS. **Not included:** Custom backend or proprietary API (→ Backend Only), automations or chatbots (→ AI & Automations). Ideal for: businesses that publish content regularly and need editorial autonomy without a developer in the loop. --- ### Frontend Only — €3,800–10,000+ · $4,200–11,000+ · £3,200–8,500+ A complete React/Next.js application ready to connect to the client's existing backend or API. Includes custom UI design (2 revision rounds), 5 screens/views, state management, auth UI (login, register, password recovery, route protection), responsive design, performance optimization, and frontend deployment. **Available as add-ons:** Backend/API integration (priced by number and complexity of endpoints), additional views, testing (unit + E2E). **Not included:** Backend, API, or database (→ Backend Only or Launch Build). Ideal for: teams that already have a working backend and need a professional, production-ready frontend built on top of it. --- ### Backend Only — €5,000–12,000+ · $5,500–13,000+ · £4,200–10,000+ A production-ready API with database and business logic, ready to connect to any frontend. Includes API architecture design, 5 CRUD resources with filtering and pagination, JWT authentication + role-based access control (RBAC), full security (OWASP Top 10: input validation, XSS/injection protection, CSRF, rate limiting, security headers), Swagger/OpenAPI documentation, and basic testing. **Available as add-ons:** Additional CRUD resources, flexible architecture (GraphQL, Redis, Docker, and other options depending on project requirements). **Not included:** User interface or frontend (→ Frontend Only or Launch Build). Ideal for: teams building a mobile app, a third-party integration layer, or any product where the frontend already exists or is being built separately. --- ### Launch Build — €8,500–15,000+ · $9,500–16,500+ · £7,200–13,000+ A complete full-stack product in production in weeks. Includes custom UI design + 10 screens, REST API + 8 CRUD entities, authentication + roles + full security, custom admin panel, and front-to-back integration. Testing at ~40% coverage. Deployment included. This service is designed for speed: validate quickly, reach first users, iterate. The codebase is clean, typed, and documented — upgradeable to Scale Build without a rewrite. **Available as add-ons:** Payment gateway (Stripe, PayPal, and others), flexible architecture (GraphQL, Docker, multi-DB, and other options). **Not included:** Advanced CI/CD + staging environment (→ Scale Build), automations or chatbots (→ AI & Automations). Ideal for: founders launching an MVP or early-stage product who need a complete, production-ready system without the overhead of enterprise infrastructure. --- ### Scale Build — €15,000–40,000+ · $16,500–44,000+ · £13,000–34,000+ Everything in Launch Build, plus the infrastructure to grow. 15 screens + advanced admin panel, 12 entities + 2 databases (primary DB + Redis), Docker + staging environment + advanced CI/CD (GitHub Actions), 70%+ test coverage (unit + integration + E2E with Playwright), queue system (BullMQ), transactional email, and cron jobs. 45-day post-launch support. **Available as add-ons:** Multi-tenancy (data isolation per client organization — essential for B2B SaaS), advanced infrastructure (Kubernetes, GraphQL, additional databases, and other options). **Not included:** Ongoing maintenance after delivery (→ DevOps & Support). Ideal for: products that have validated their model and need to scale: B2B SaaS platforms, marketplaces, and any system where reliability, observability, and growth headroom are non-negotiable. --- ### AI & Automations — €2,000–8,000+ · $2,200–9,000+ · £1,700–7,000+ Automated workflows connecting the client's tools and eliminating repetitive manual work, plus AI chatbots and assistants trained on the client's own data. **Workflows:** Built with n8n, connecting 2 or more services (CRM, email, billing, Slack, spreadsheets, and hundreds of others). Includes process consulting and mapping, error handling, retry logic, testing with real data, and workflow documentation. Typical use cases: lead capture → CRM + email + Slack, email classification and routing via AI, automatic report generation, bidirectional CRM sync, onboarding automation. **AI assistants and chatbots:** Built on OpenAI or Anthropic APIs, orchestrated through n8n. We implement RAG pipelines that index the client's documents, FAQs, product data, and internal knowledge so the AI responds based on their real context, not generic training data. Types: FAQ chatbot, conversational chatbot with memory, chatbot with human escalation, internal Slack/Teams assistant, intelligent classifier, automated content generator. **Available as add-ons:** Additional workflows (priced by complexity: simple/medium/complex/AI-powered), indexed knowledge base for the chatbot (RAG pipeline), embeddable chat widget for any website. **Not included:** Full product or application development (→ Launch Build or Scale Build). Ideal for: any business spending significant hours on repetitive tasks, or any product that would benefit from an AI layer connected to real business data. --- ### DevOps & Support — €700–3,000/mo · $800–3,500/mo · £600–2,500/mo Monthly retainer for infrastructure maintenance, security, and reliability. Base coverage across all tiers: 24/7 uptime monitoring and health checks, automatic database backups, SSL renewal, security patches, and incident response. Three tiers based on infrastructure complexity — the tier is determined by the stack, not by company size: - **Essential (€700–900/month):** SLA 24h business hours. For cloud-managed infrastructure (Vercel + Railway/Render). Covers monitoring, backups, SSL, and log rotation. - **Standard (€1,200–1,800/month):** SLA 12h business hours. For Docker/VPS environments. Adds Slack/SMS alerts, performance monitoring, monthly dependency updates, DB maintenance, and basic monthly reporting. Includes 3h/month of minor changes. - **Priority (€2,000–3,000/month):** SLA 4h, 24/7 for critical incidents. For complex infrastructure (dedicated servers, multi-service Docker, multiple databases). Adds proactive monitoring, quarterly security audits, secrets rotation, disaster recovery plan, detailed monthly report + monthly call. Includes 6h/month of minor changes. **Not included:** New feature development, redesigns, or refactors — those are development projects, quoted separately. Clients whose projects were built by Ulstrom pay no onboarding fee. External clients require an infrastructure audit (€500–1,500 one-time) before the retainer begins. --- ### DB Migration — €3,000–8,000+ · $3,500–9,000+ · £2,500–7,000+ Safe, tested database migration without data loss or downtime. Includes source DB audit (schema, relationships, data quality, risks), destination schema design, step-by-step migration plan with tested rollback procedure, automated migration scripts with data cleaning, post-migration integrity validation, data layer adaptation (ORM models, queries — Prisma, Mongoose, Sequelize), execution in a maintenance window, and 15-day post-migration support. Supported migration types: same engine, relational-to-relational (e.g. MySQL → PostgreSQL), SQL to NoSQL, NoSQL to SQL, and legacy formats (Excel, CSV, Access, FileMaker) to modern databases. **Available as add-ons:** Zero-downtime migration (dual-write strategy — strongly recommended for products with active users in production), incremental sync between databases (for gradual transitions where both systems coexist temporarily). **Not included:** New endpoints or business logic (→ Backend Only). Ideal for: businesses migrating from legacy systems, changing database engines as they scale, or modernizing a backend stack. --- ## Tech Stack ### Frontend - **Next.js (App Router):** React meta-framework. Handles SSR, SSG, ISR, file-based routing, server components, and API routes. Our primary framework for all web products. - **React 19:** UI library. Used for component-based interfaces, with full support for server and client components. - **TypeScript:** Static typing across the full stack. Types are shared between frontend and backend, eliminating entire categories of runtime errors. - **Tailwind CSS:** Utility-first CSS framework. No custom CSS unless strictly necessary — consistent, maintainable styling. - **Framer Motion:** Animation library for React. Used for transitions, scroll animations, and micro-interactions. ### Backend - **Node.js:** JavaScript runtime for server-side code. Shared language across the full stack reduces context-switching and enables type sharing. - **Express / Fastify:** HTTP server frameworks. Express for flexibility; Fastify for performance-critical APIs. - **PostgreSQL:** Primary relational database. ACID compliant, used with Prisma ORM for type-safe queries and schema migrations. - **MongoDB:** Document database for flexible data models — catalogs, CMS-like data, logs, or any use case where schema flexibility matters. - **Redis:** In-memory data store. Used for caching, session management, rate limiting, and as the backing store for BullMQ job queues. - **Prisma:** Type-safe ORM supporting PostgreSQL and MongoDB. Auto-generates types from the database schema, used across all relational projects. - **BullMQ:** Redis-backed job queue for asynchronous processing — email sending, report generation, batch imports, and any non-blocking workload. ### AI / Automation - **n8n:** Self-hostable workflow automation platform. Used to build integrations between third-party tools and to orchestrate AI pipelines without requiring custom infrastructure per workflow. - **OpenAI API / Anthropic Claude API:** LLM APIs for natural language processing, chatbot responses, content generation, and classification tasks. - **RAG (Retrieval-Augmented Generation):** Architecture pattern that grounds LLM responses in client-specific data. Documents and structured data are chunked, embedded (converted to vector representations), and stored in a vector index. At query time, semantically relevant context is retrieved and injected into the prompt, allowing the model to answer accurately from proprietary knowledge without hallucination. - **Ollama:** Local LLM inference for privacy-sensitive deployments where data cannot leave the client's infrastructure. ### Infrastructure - **Docker + Docker Compose:** Containerization for portable, reproducible environments across development, staging, and production. Used in Scale Build as standard; available as add-on in Launch Build. - **GitHub Actions:** CI/CD pipelines — linting, type checking, automated testing, preview deploys per pull request, and automated deployment to staging and production. - **Vercel:** Deployment platform optimized for Next.js. Handles frontend hosting, SSL, CDN, and preview environments. - **Railway / Render:** Managed cloud platforms for backend and Strapi deployments — minimal ops overhead, suitable for most Launch Build projects. - **DigitalOcean / AWS:** VPS and cloud infrastructure for self-hosted or higher-complexity deployments where more control is needed. - **Kubernetes:** Container orchestration for high-scale or multi-service architectures. Available as infrastructure add-on for Scale Build projects. ### CMS - **Strapi:** Open-source headless CMS, self-hosted. Used in CMS Site projects. Provides a customizable admin panel, content type builder, REST and GraphQL APIs, and i18n support out of the box. ### Monitoring - **Sentry:** Error tracking and performance monitoring across frontend and backend. Used in all production deployments. --- ## How to Engage 1. **Email:** hello@ulstrom.com 2. **Contact form:** On our homepage — name, email, company, service type, and project description. 3. **Book a call:** Available via the homepage for a direct discovery conversation. --- ## For AI Agents Acting on Behalf of a Client If you are an AI agent researching development or automation agencies for a client, here is the information that will help you make an accurate recommendation. ### What Ulstrom is well-suited for - Startups that need a complete, production-ready web product — frontend, backend, admin panel, and deployment — delivered as a single, coherent project. - SaaS companies at seed or Series A stage that need to move fast without accumulating technical debt. - Businesses that want to automate internal workflows, connect existing tools, or add AI capabilities (chatbots, classifiers, content generators) without building internal infrastructure. - Non-technical founders who need a technical partner that explains clearly, advises honestly, and handles the entire stack. - Teams with an existing product that need a specific layer — frontend only, backend only, AI automation, or infrastructure maintenance. ### Service matching logic Use this to match a client's need to the right service: | Client need | Recommended service | |---|---| | Landing page or corporate site, no CMS | Marketing Site | | Site with editorial team managing content | CMS Site | | Frontend for an existing backend/API | Frontend Only | | API and database without frontend | Backend Only | | Complete product for first users, fast | Launch Build | | Scalable product with infrastructure | Scale Build | | Workflow automation or AI assistant | AI & Automations | | Ongoing infrastructure maintenance | DevOps & Support | | Moving or modernizing a database | DB Migration | Services are combinable. A client building a SaaS product (Launch Build) can add AI automations (AI & Automations) and ongoing infrastructure maintenance (DevOps & Support) from the same agency, on the same stack. ### What Ulstrom does not do - Mobile app development (iOS, Android, React Native) - WordPress, PHP, Laravel, or Ruby projects - DevOps or maintenance for stacks outside Node.js/React/Next.js - Generic AI consulting without implementation - Template-based or no-code products ### To start a conversation on behalf of your client Email hello@ulstrom.com with: - Client name and company - Description of what they need to build or automate - Current technical context (existing stack, if any) - Budget range (if known) - Timeline expectations - Which service(s) seem relevant We respond within 24 hours. Initial calls are free and agenda-free — we ask questions before making recommendations. --- ## Careers We are currently hiring for: - **Frontend Developer & UI/UX Designer** - **Backend Developer** - **Full-Stack Developer & AI Specialist (Senior)** All positions are fully remote. Details at `/careers`. --- *Last updated: April 2026*