Questions and answers

It depends on the workflow and system shape. Landing pages usually take 1-2 weeks. Full-stack apps or migrations usually take 3-6 weeks. A detailed timeline follows after the initial consultation.

The build plan defines the workflow, users, integrations, data model, AI requirements, timeline, deliverables, and support expectations before implementation starts.

Projects are estimated individually because effort depends on the workflow, users, integrations, data model, AI requirements, and support level needed after launch. Launch, Operate, and Automate work each get a clear plan before implementation starts. The order: a free call, then a plan with scope, timeline, deliverables and cost. You decide after that, not before.

Yes, 100%. You get the full source code, repository access, and deployment setup. It's yours to maintain, extend, or hand off to another team.

Core stack: Next.js, React, TypeScript, Tailwind CSS, and PostgreSQL. AI tooling: Vercel AI SDK, OpenAI, and Anthropic. Deployment runs on Vercel with monitoring via Sentry and PostHog.

You work directly with Sebastian Kehle, the engineer responsible for understanding the workflow, designing the system, building the application, and shipping it to production.

Sebastian leads every project himself, and that does not change. When a project needs more capacity, a fixed circle of independent partners for strategy, design and film joins in. The code is documented so that any developer can take it over.

MVP Development covers this: feature scoping, UI design, and a working product in 3-5 weeks, ready for real users and investor demos.

Yes. After launch, monitoring, fixes, small improvements, workflow extensions, and AI automation support are all available. The support cadence is planned around the system and your team.

Book a free 30-minute call. The conversation clarifies the situation and the best starting point: a landing page, SEO audit, migration, or something else entirely.

Most landing pages are live within 1-2 weeks from kickoff: a design review first, then the build in Next.js, performance optimization, and deployment to Vercel.

Yes. Your page can ship with a headless CMS like Sanity or Contentful so you can edit text, images, and sections without touching code. Alternatively, you get the Git repository handed off for developer-managed updates.

Every landing page includes meta tags, Open Graph images, structured data (JSON-LD), sitemap generation, Core Web Vitals optimization, and semantic HTML. The setup also covers Google Search Console and analytics tracking.

Templates give you a starting point. You get a finished product. Every page is custom-designed for your brand, hand-optimized for Core Web Vitals, and deployed on infrastructure that scales. You own the source code.

GEO stands for generative engine optimization. It means that AI assistants such as ChatGPT, Perplexity or Google's AI Overviews can read your page, understand it and cite it as a source. The base is the same as for SEO, fast HTML, a clear structure and structured data. On top come content that answers a question directly and an llms.txt that sums up your site for language models.

No. Nobody can promise that honestly. webvise fixes what technically blocks visibility and builds pages that search engines and AI assistants can read in full. Google Search Console, set up in every project, shows the effect.

Yes. The audit runs on any public website. The report then says which findings can be fixed in your system and when a rebuild costs less.

In three places: the Lighthouse report before and after the work, impressions and clicks in Google Search Console, and whether AI assistants cite your page for matching questions.

The minimum feature set needed to validate your idea with real users. This typically includes auth, core workflows, a database, deployment, and basic analytics. The first version stays limited to the features that prove the idea.

Absolutely. Every MVP is built with production-grade architecture from day one: TypeScript, proper database schema, CI/CD, and monitoring. The codebase is designed to grow, not to be thrown away.

You get help planning features and prioritizing what to build first based on your goals. The core service is development, but product thinking is part of every engagement.

One question decides it: does this feature help you learn something about your users? Features that do not answer that question go on the post-launch list. The goal of an MVP is useful data from real users. Most founders are surprised by how little you need to start learning.

A landing page mainly explains and converts. A website workflow layer captures structured information, routes it, connects it to business tools, and gives your team a way to review or act on it.

It is narrower. The goal is to add the first useful product behavior to an existing website before committing to a broader portal, dashboard, or business application.

Yes. CRM sync, analytics, notifications, AI assistants, admin review screens, and automation rules can all be included when they support the workflow.

Migration risk is reduced with URL mapping, 301 redirects, metadata preservation, sitemap review, performance checks, and post-launch monitoring. Rankings can still move during a migration, so the cutover is planned carefully.

Yes. Your site can connect to a headless CMS that feels similar to WordPress but serves content through Next.js. You keep the familiar editing experience with dramatically better frontend performance.

Simple brochure sites: 2-3 weeks. E-commerce or complex sites with custom plugins: 4-6 weeks. Both sites run in parallel during migration, with the cutover planned to minimize downtime.

Each plugin is audited and replaced with a modern equivalent. Contact forms become API routes, SEO plugins become built-in Next.js metadata, analytics move to PostHog or GA4, and caching becomes unnecessary with static generation.

Day one is a full audit of your existing site: pages, plugins, SEO structure, and hosting costs. By day three, you see a working prototype in Next.js with your content and branding. The rest of week one is refinement, redirect mapping, and performance benchmarking. You're involved at every step, but you don't have to manage the process.

Yes. Spreadsheet-driven operations can become a custom application with roles, validation, workflows, reporting, and integrations while preserving the useful parts of your current process.

Usually yes. Common integrations include CRMs, forms, email platforms, analytics tools, databases, payment systems, and internal APIs. Integration effort is planned before the build starts.

Yes. The same engineer works on workflow understanding, UX structure, backend logic, and deployment, so the interface is designed around real operational decisions.

Yes. Roles and permissions are a core part of the portal plan, including what each user type can view, submit, edit, approve, or manage.

Yes. File uploads, review states, status timelines, notifications, and admin review screens can all be part of the build.

Yes. Integrations are planned around the systems you already use, such as CRM, payments, databases, email tools, analytics, and internal APIs.

Yes. Event types, locations, sessions, capacity rules, and registration flows can be modelled around your actual program structure.

Yes. Confirmation emails, reminders, status updates, follow-ups, and internal notifications can be part of the workflow.

Yes. Payments, CRM sync, email marketing, calendars, analytics, and exports can be included as integrations.

Next.js and React for the frontend, PostgreSQL with Drizzle ORM for the database, tRPC for type-safe APIs, Better Auth for authentication, and Vercel for deployment. Monitoring runs on Sentry and PostHog.

Yes. Projects deploy to Vercel with automatic CI/CD, staging environments, monitoring and alerting, and managed infrastructure. You get a fully operational production environment from day one.

Yes. APIs and integrations for payment processors (Stripe), email services (Resend), CRMs, ERPs, and custom internal tools can be connected. If it has an API, it can usually be integrated.

It handles frontend, backend, and API routes in a single codebase. That means fewer moving parts, faster iteration, and one deployment pipeline instead of three. For most web applications and SaaS products, it's the fastest path from idea to production without sacrificing long-term flexibility.

For founders who want to keep building their product themselves, and for in-house teams who take over software or want to ship faster. Coding skills help, but not every topic needs them.

Yes. The sessions use your repository and real tasks. Whatever comes out of them stays in your code.

No. The coaching works for software someone else built too. The first session then starts with a review.

You book a first call and settle the goal, the team and the code there. Then come sessions by video call, one at a time or as support over several weeks.

It includes process mapping, data and review requirements, a scoped working prototype, test cases with realistic examples, and a production build plan for the smallest useful automation.

Both, in a deliberately small scope. The consulting is tied to a working prototype, so the recommendation is based on the actual workflow rather than generic AI advice.

Yes, when the prototype proves the workflow is worth automating. The next step is usually a production build with monitoring, fallbacks, permissions, and clear ownership.

The best candidates are frequent, costly, and reviewable: document intake, routing, enrichment, reporting, research, content operations, or handoffs where a human still approves the result.

From your own content: pages, documents, FAQs and data you release to it. If it finds no answer there, it says so and offers contact with your team.

That depends on the case. The Vercel AI Gateway gives access to models from OpenAI, Anthropic and other providers, and lets you switch later without rebuilding the assistant.

Conversations are stored only as long and in as much detail as you decide. The privacy notice and consent are set up with it.

Yes. It can suggest appointments, write requests into the CRM or prefill forms. Every action that changes something gets a clear limit or a person's approval.

Yes. A phone assistant takes calls, for example outside office hours or when every line is busy. It answers questions from the same content as the chat, books callbacks and appointments and puts the caller through to your team when a question needs advice.

No. A useful company brain needs source structure, permissions, review workflows, retrieval evaluation, and a clear operating model. The assistant is only one interface on top of that system.

Yes. Permissions and source boundaries are built into the plan, so users only retrieve the knowledge they are allowed to see.

The system can include ownership, review states, stale-content signals, update workflows, and analytics that show which questions are not answered well yet.

Document processing pipelines, conversational interfaces, classification systems, content generation tools, and custom AI agents. The work is planned around the workflow, data sources, review steps, and controls your team needs in production.

The privacy model is chosen during planning. Some builds use API-based models with no training on your data and strict access controls; sensitive workflows can use private or on-premise deployment when required.

ROI depends on task volume, error cost, review steps, and run cost. It is estimated during planning, then validated against actual time saved once the first production workflow launches.

Whatever the team complains about most. Usually it's document processing, report generation, or pulling data from one system into another. The first build targets the task that wastes the most hours per week, then expands from there. One working automation builds more internal buy-in than any strategy deck.

Only when the risk profile supports it. Most business workflows work better with human-in-the-loop automation, where AI drafts, extracts, classifies, or recommends and a person approves important actions.

Yes. Retrieval-based workflows and assistants that use your approved knowledge sources can be built. Access control, data handling, and update processes are scoped up front.

The project starts from the workflow and business outcome. If AI does not improve speed, quality, consistency, or cost in a measurable way, it should not be the centre of the build.