Senior Next.js Developer (AI/LLM) — Project-Based, Paid Per Feature — Work Health & Safety SaaS Platform

About the Role
We are looking for a Senior Next.js Developer with strong, hands-on AI experience to deliver features on a project basis for a production-grade, multi-tenant SaaS platform for work health and safety compliance management, used by Australian businesses in construction, warehousing and manufacturing.

Work is delivered feature by feature, ticket by ticket. Each ticket has a clear scope, acceptance criteria and a fixed price agreed before you start. There is an ongoing pipeline of work for the right developer. This is not a beginner role.

How We Work

Work is issued as Jira tickets with Figma designs and acceptance criteria
You quote or accept a fixed price per ticket before starting
You deliver via pull request to a preview environment
Our lead developer reviews the code; once accepted, the ticket is paid
Payment via OnlineJobs.ph EasyPay

What You Will Do

AI Feature Development — Design, build and ship production-ready AI features (e.g. document and policy drafting, risk assessment assistance, inspection and hazard analysis, smart search across procedures, compliance insights and reporting)
LLM Integration — Work with LLM APIs (Anthropic Claude, OpenAI or similar), including prompt design, structured outputs, tool/function calling and streaming responses
RAG & Search — Build retrieval pipelines over client documents (embeddings, vector search with pgvector or similar, chunking, evaluation)
Multi-tenant Safety — Ensure AI features respect tenant data isolation, permissions and privacy at all times
Reporting & Exports — Build report exports (PDF, Excel, CSV) and dashboards
App Development — Build and extend features across our Next.js frontend and backend from Figma designs
Bug Fixing & Refactoring — Improve existing Server Actions and components without breaking unrelated features

Tech Stack (Must Have)

Next.js 15 (App Router) — Server Actions used extensively
TypeScript (Strict) — No any types; type safety is a priority
Styling — Tailwind CSS, Radix UI, Shadcn UI
Database — PostgreSQL with Kysely query builder (via Neon)
State & Forms — React Context, React Hook Form with Zod validation
AI — Proven experience shipping LLM features to production (not just tutorials or side projects); Vercel AI SDK or similar

Bonus Skills

pgvector / embeddings / RAG evaluation
AI agents and tool calling / MCP
Document processing and generation (PDF/DOCX/XLSX, OCR)
AWS S3 uploads via presigned URLs
Replicache or local-first sync strategies
Experience in compliance, safety, HR or other regulated SaaS

Who You Are

You have shipped real AI features that real users rely on, and can show us
You estimate accurately and deliver what you quote, on time
You pay attention to detail — if the design has 16px padding, you don't make it 20px
You're comfortable working in a large, existing codebase
You communicate clearly in English and flag issues early
You can overlap with Australian business hours (AEST/AEDT) for check-ins

Selection Process
Shortlisted candidates will complete a paid first ticket: building report export options (PDF, Excel, CSV) with an AI-generated executive summary, using mock data. Developers who deliver well move on to ongoing tickets.

How to Apply
To confirm you've read this post, please do the following:

Use the subject line: "Next.js AI Developer Application - [Your Name]"
Include a link to your portfolio or GitHub profile
Describe one AI feature you have shipped to production: what it did, the model/API used, and how you handled cost, accuracy or hallucinations
Answer this question: "How would you stop one client's documents from appearing in another client's AI search results in a multi-tenant app?"
Tell us how many hours per week you can commit and your usual hourly rate