Brilad builds websites and integrations on WordPress, Next.js and Laravel that search engines and AI crawlers can read in full, and that connect business data to AI assistants through the Model Context Protocol (MCP). Many visibility problems come from how a site is built, such as content that only appears after JavaScript runs or schema that conflicts between plugins. Fixing them in the code is often faster than working around them in content.
Development work is led by co-founder Ashruba Pankhade, who has five years of experience building web applications with Laravel and PHP, including APIs and payment integrations.
AI-ready WordPress
This website is our own example. It runs on a custom WordPress theme we wrote, with these choices built in:
- Fonts served from the site itself, no jQuery on the front end, and scripts loaded with defer
- One JSON-LD graph for the organisation, website and articles, which switches itself off when Yoast, Rank Math or a similar plugin is active so the page never carries two competing graphs
- An llms.txt file generated from the site’s service pages and articles, cached and refreshed when content changes
- Optional robots.txt rules that block AI training crawlers while leaving AI search crawlers allowed
- Hardening at theme level: XML-RPC disabled, user lookups through ?author= and the users REST endpoint blocked for visitors, and the admin file editor switched off
For client sites we either build a theme like this or fix the existing one: removing page-builder bloat from templates, correcting schema, and moving content out of sliders and tabs that load late.
WordPress 6.9 added the Abilities API, which lets plugins declare actions that AI agents can call through an MCP adapter. We set up these connections with a dedicated user on a limited role, OAuth sign-in, an activity log, and new content saved as drafts, so an AI assistant can help with content without being able to change plugins, themes or settings.
Next.js development
Next.js can render a page on the server for each request, build it ahead of time, or regenerate it on a schedule. The choice matters for crawlers. A page rendered only in the browser sends little more than an empty shell to any crawler that does not run JavaScript, and most AI crawlers do not.
On Next.js projects we:
- choose server rendering or static generation for every page that should be found in search or cited by AI
- generate titles, descriptions and canonical URLs with generateMetadata from the same data that builds the page
- produce sitemap.xml and robots.txt from code (the sitemap.ts and robots.ts file conventions), so they stay in step with the actual routes
- output JSON-LD from the page data instead of hard-coding it
- connect a headless CMS, often WordPress, when editors need a familiar interface
Laravel and AI integrations
MCP is an open protocol, released by Anthropic in November 2024, that lets AI assistants call tools and read data from other systems. An MCP server built into your Laravel application can let an assistant search your product catalogue, draft content in your CMS or look up order status, within the permissions you define. Laravel has an official MCP package for building these servers.
The security model matters more than the feature list. We build MCP servers with OAuth-based sign-in, per-tool permission checks, rate limits and an audit log of every call. Tools that change data are kept separate from read-only tools, so read access can be granted without write access.
We also build AI features inside applications, such as search that understands natural-language questions and chat that answers only from your own documents, with links back to the source page for each answer.
Stack
| Layer | What we use |
|---|---|
| CMS and frontend | WordPress (classic and block themes), Next.js, React |
| Backend | Laravel, REST APIs |
| AI connectivity | Model Context Protocol, WordPress Abilities API, llms.txt |
| Hosting | AWS, cPanel and WHM servers |
| Structured data | JSON-LD with schema.org vocabulary |
Frequently asked questions
Do we need a new website to rank in AI answers?
Usually not. Most sites can be fixed with changes to templates, rendering and schema. A rebuild is worth considering when the platform cannot serve content without JavaScript or when the theme cannot be changed safely.
Is it safe to connect an AI assistant to our website?
It can be, if the assistant signs in as a user with limited permissions, every action is logged, and nothing it writes goes live without human review. An assistant connected with an administrator account or a shared API key is a real risk, and we do not set up connections that way.
Can you work with our existing developers?
Yes. We can review their code, build specific parts alongside them, or hand over documented code for them to maintain.
Further reading
- What is llms.txt and does it help AI search?
- JSON-LD schema for AI search
- AI crawlers and robots.txt
- Retrieval-augmented generation
- Machine readability and clean HTML
- IndexNow: faster indexing for Bing and AI search
Start with an AI Search Audit to find out what your site needs.