LLM Engineering Intern

Corvio is building an AI-native doc system from zero to one.

We believe the important problem is no longer just helping AI answer questions or write text. The real challenge is turning AI conversations and scattered information into a system that can keep organizing, updating, and evolving over time.

Today, valuable work is still scattered across chats, docs, meetings, files, screenshots, emails, and feedback loops. Traditional doc systems were built for manual upkeep, and traditional AI chat was built for one-off prompting. Corvio is rebuilding that layer so structure, memory, judgment, and agent collaboration can operate as one continuous system that becomes clearer, more aligned, and more useful with use.

We are an early, small, high-density team based primarily in Singapore. We are looking for people who want to help define this system with us, not just complete a conventional internship track.

You will work on questions such as how agents keep operating in real document environments instead of stopping at one-shot calls, how long-term context should be structured, and how model behavior can remain stable, controllable, and traceable inside a complex product system.

This is not a role for patching edge features. You will be contributing to a broader foundation for how humans and AI should work together inside one evolving system.

• LLM orchestration and agent runtime.
• RAG, long-term memory, and context management.
• Deep integration between AI capabilities and the document system itself.
• Structured knowledge, execution chains, and state system design.
• Coordination across prompt, tool, memory, and retrieval layers for real product scenarios.
• Eval, stability, and iteration work against a live environment rather than demo-only prototypes.

• Currently enrolled in an undergraduate program or above.
• Strong interest in AI, programming, and system design, with your own technical judgment.
• Solid Python fundamentals and good engineering habits.
• Practical understanding of LLMs, RAG, agents, or context engineering through projects or hands-on work.
• Enjoyment of difficult problems that still lack consensus answers.
• Ability to learn quickly, push independently, and keep moving toward first-principles understanding in ambiguous settings.
• Ownership over outcomes rather than only finishing assigned work.
• Availability for 4 or more days per week over at least 4 months.

• Experience building AI Native products, agents, RAG systems, search, recommendation, or other complex application systems.
• Hands-on experience with prompt orchestration, tool use, memory, eval, or retrieval pipelines.
• Backend, infrastructure, or full-stack engineering experience.
• Long-term interest in knowledge management, document systems, collaboration products, or workflow products.
• Heavy practical use of tools such as ChatGPT, Claude, and Cursor in real projects.

• Direct exposure to the most central and forward-looking problems inside an AI Native product.
• Hands-on work with real LLM engineering in a live system instead of demo-style development.
• Frequent collaboration with founders and core teammates in a high-density environment.
• A chance to understand the real tension between model capability, systems engineering, and product reality.
• An experience closer to defining a new product category than completing a standard internship.
• Potential long-term continuation, full-time conversion, and startup equity incentives.

• 4 or more days per week.
• 4 or more months.
• Primarily remote collaboration.