Front-End Engineer - Website Platform

About MyAdvice
MyAdvice is an established, profitable company with a long history in the markets we serve. We have made a fast, deep shift toward AI-enabled solutions and operate with the urgency and standards of a startup. We are fast-moving and aggressive. We need results-oriented people on our team who will never be satisfied with the status quo and will work hard toward continuous improvement.

What we do. MyAdvice provides an AI-powered system that reduces manual work and cost while helping clients grow revenue. The Success Center, powered by Mya, unifies growth activity across channels and automates workflow execution.

Our mission. We make it easy to profitably grow your business.

How we work. GREAT means Gritty, Resourceful, Expert, Accountable, and Tenacious. We expect hard work, extreme ownership, and extreme accountability. We are willing to do the difficult work required to deliver exceptional outcomes for our clients and our colleagues.“We are going to relentlessly chase perfection, knowing full well we will not catch it, because nothing is perfect. But we are going to chase it because, in the process, we will catch excellence.” — Vince Lombardi

The opportunity
We're hiring the front-end owner of a new website platform: the engineer responsible for everything a client, a marketer and a patient actually see and touch.

MyAdvice powers the websites of thousands of medical, dental and legal practices. Today those sites live on a fragmented legacy stack, each one built and patched by hand. We're replacing all of it with one platform built on Next.js and Payload, where a new client site is provisioned in minutes, and we're building a generative AI pipeline that will design, lay out and populate it.

The back end of that platform has owners: multi-tenancy, the data model, provisioning and infrastructure. What it needs now is someone who owns the experience layer as seriously as they do. That means the editor people build pages in, the component library every site is assembled from, the theming system that makes each site look bespoke from one codebase, and the rendering and performance work that decides whether those sites get found.

Two things make this job unusual, and you should know both before you apply.

We are building our own editor, not configuring someone else's. Payload's admin is built for developers, and there is no client-ready page builder we can drop in. We evaluated the off-the-shelf options and decided against them. So the editor is ours to build, over Payload's native blocks schema, working inside the framework's admin internals. This is the deep end of Payload, and it is the hardest and most consequential front-end work on the platform. Thousands of practice staff who use Elementor, Divi or Gutenberg today will land in what you build.

The AI generates configuration, never code. One framework, one component set, unlimited variation through data. The pipeline emits schema-validated JSON: a token set, a section composition, per-block style properties. Your component library and token system are the vocabulary that output is written in, which means the quality ceiling of every generated site is set by what you build.

Every improvement you make ships to every site at once. A 200ms LCP win is not one page, it is the whole fleet.

You'd join a small, Website Platform team: a platform engineer on the application and infrastructure layer, a full-stack builder, a senior architect who knows our existing systems deeply and advises across the project, and our AI engineering group on the generation pipeline. You'd report to our VP of Engineering.

What you'll own

• The editor. Our own editing experience, built over Payload's native blocks schema, that non-technical marketers and delivery staff can move fast in without being able to break a site. Drag and drop, live preview, undo, validation, and guardrails that hold. This is product engineering inside a framework's internals, and it is the centre of this role.
• The component library. The blocks every site is assembled from: heroes, provider directories, location maps, before-and-after galleries, review carousels, forms. Typed, accessible, tested, documented, and composable enough that the AI pipeline can assemble them as reliably as a person can.
• The theming system. One codebase, per-tenant design tokens, genuinely distinct sites, zero per-tenant code. You'll own the token vocabulary and the mechanism that applies it. You'll also hold the line that keeps it working: tokens rather than per-block overrides, no escape hatch for custom CSS, new variants earning their place through review rather than accumulating.
• Rendering and performance. The Next.js App Router rendering strategy, covering server components, static generation, caching and revalidation, and Core Web Vitals across the whole fleet. You measure before you claim, and you set the performance budgets that keep new work from regressing.
• SEO in the markup. Semantic HTML, metadata, and valid schema.org structured data for practices, providers and locations. These sites exist to be found.
• Accessibility. WCAG 2.1 AA as a floor, not an aspiration. Healthcare websites serve patients of every ability, accessibility failures are a legal risk for our clients, and contrast and a11y checks are a gate that generated output has to pass before it ships.
• Rendering what the AI generates. The pipeline produces a layout and content as data. Your components and editor turn that into a site a practice will put its name on. You'll work with our AI engineers on the contract between the two, and make sure a generated page can always be opened, understood and fixed by a person.
• Handling patient media correctly. Before-and-after imagery is protected health information, and it flows through the upload path and gallery components you build. You'll be responsible for getting consent state, access control and provenance right in the front end, with our architect and compliance work behind you.
• The front-end quality bar. Testing strategy covering unit, component, visual regression and end-to-end, front-end CI checks, code review standards, and the patterns the rest of the team follows.

What you'll work within
Two constraints are worth stating plainly, because they shape the work and we would rather you know now.
The page schema is frozen early and deliberately. It is the one commitment on this platform that is expensive to reverse, so it gets locked in the first phase and changes after that are rare and deliberate. You'll be building a large component library inside a schema you cannot casually reshape.
We also expect to work with an external build partner during the build phase, who will contribute code and then exit. Part of this role is reviewing what they produce and owning it afterwards, which is a real responsibility and also a fast way to see the whole surface of the platform.

What we're looking for

• We care about what you have shipped and can show us, not what is listed in your skills section. Specifically:
• 5+ years building production front ends, with React and TypeScript as your daily tools. You've shipped and maintained real applications, not prototypes.
• Next.js in production, ideally the App Router. You can explain when you'd render on the server, statically or on the client, and you've lived with the caching consequences.
• You've built custom editing or admin UI inside a framework you didn't write. Payload admin components, Sanity Studio customisation, Strapi admin plugins, a custom canvas over a CMS data model, or comparable depth elsewhere. We want evidence you can work inside someone else's framework internals rather than only on top of a clean API. This is the requirement we weigh most heavily.
• You've built a component library or design system that other engineers used. You've designed component APIs, versioned them, documented them, and dealt with the requests to bend them.
• You've rendered UI from data. Server-driven UI, a block or schema-based renderer, layouts described as JSON rather than written as JSX. You understand what makes a component set composable by a machine rather than only by a person.
• Measured performance work. You've moved a Core Web Vital, whether LCP, INP or CLS, and can tell us the before, the after, and what you changed.
• Accessibility you can demonstrate. Keyboard navigation, focus management, ARIA used correctly, and testing with real assistive technology, not just a passing Lighthouse score.
• You test your front end. Component tests with Jest or Vitest and Testing Library, plus end-to-end or visual regression tests with Playwright, Cypress or similar, are part of how you ship.
• Enough back end to be dangerous. You're comfortable reading and writing Node.js and working against a headless CMS or API, even if the server isn't your main job.
• You raise the bar around you. Code review, patterns, documentation, mentoring.
• AI-native. You use Cursor, Claude or similar to build faster, and you own the judgment about what is safe to ship to a live fleet.
• Overlap with US colleagues. Comfortable working 2pm to 11pm IST with real-time collaboration.