AI Software Engineering Intern

Cleve; Kuala Lumpur; On-site; MYR 100,000 - 145,000; Full time

Job summary

Cleve is hiring a full-time engineer to ship full-stack features across a Next.js frontend and Convex backend. You’ll build context-aware AI pipelines, evaluate LLM providers, and work with embeddings and vector search to retrieve knowledge.

You’ll collaborate with the CTO and team from idea to deployment, shipping early v1s and iterating quickly in a fast-paced AI-Driven environment.

Qualifications

• Proven ability to ship features to real users.
• Strong JavaScript/TypeScript skills and Git.
• Hands-on experience with LLMs and embeddings.

Responsibilities

• Ship full-stack features across Next.js frontend and Convex backend.
• Build and improve RAG pipelines for context-aware AI responses.
• Integrate and evaluate LLM providers like OpenAI, Anthropic, and Gemini.
• Work with embeddings, vector search, and knowledge retrieval.
• Experiment with prompts, evals, and agent behaviours.
• Collaborate with the CTO and team from idea to deployment.

Skills

JavaScript/TypeScript

Git

LLMs

AI tooling

Tools

Next.js

Convex

Serverless

Job description

Ship production code at the intersection of software and AI, to more than 40,000 real users. This is not a fetch-coffee internship. You'll own features across the stack, work directly with our CTO, and push to production in your first week. We move faster than companies ten times our size, and we're looking for people who already build like that.

What you'll be doing

• Ship full-stack features across our Next.js frontend and Convex backend
• Build and improve RAG pipelines for context-aware AI responses
• Integrate and evaluate LLM providers like OpenAI, Anthropic, and Gemini
• Work on embeddings, vector search, and knowledge retrieval
• Experiment with prompts, evals, and agent behaviours
• Work with the CTO and team from idea to deployment

What we're looking for

• A GitHub or portfolio with things you've actually built and shipped. Side projects, hackathon entries, apps with real users. The bar to ship has never been lower, and we expect you to have used that to your advantage
• Proficiency in JavaScript/TypeScript and Git
• A working understanding of LLMs and embeddings from using them, not just reading about them
• You build with AI tools daily and have opinions about which ones are good

Nice to have

• A project with real users and the scars to show for it
• Experience with Next.js, Convex, or serverless architecture
• Familiarity with vector databases and retrieval systems
• Open source contributions

We're attracted to someone who

• Builds things before anyone asks them to
• Would rather ship a rough v1 today than a polished v1 next month
• Keeps up with AI by trying things, not just reading threads
• Takes initiative rather than waiting for instructions
• Can unlearn and relearn quickly in a field that changes monthly