AI Product Engineer Intern

About AugvantAugvant is an early-stage, AI-native startup based in Taipei. We are redesigning enterprise workflows from the ground up—building systems where AI doesn’t just answer questions, but actively executes and participates in core business operations.Backed by our paid enterprise design partner, we are transforming how traditional accounting firms handle billing reconciliation, document processing, tax filing, and internal operations—turning fragmented manual processes into verifiable, auditable software systems.We believe the next generation of software engineers aren’t just people who write syntax. They are builders who leverage AI to drive an idea all the way from ambiguous problem definition to reliable production delivery.*Augvant is a member of the Google for Startups Cloud Program.What You’ll DoYou will work directly with the Founder to build and ship real enterprise software used in production:Deconstruct Problems: Dive into real user workflows and edge cases to translate messy, ambiguous operational needs into clear functional specs and acceptance criteria.Build with AI: Leverage Coding Agents (Claude Code, Codex, Cursor, etc.) to build full-stack features, APIs, data pipelines, and automated workflows at high velocity.Verify & Harden Quality: Write tests, hunt down edge cases, and debug systematically—ensuring features don’t just "work on my machine," but run reliably in production.Ship Clean Code: Use Git, PRs, and Code Reviews to deliver maintainable, auditable codebases.Iterate in the Real World: Continuously refine the product based on live user feedback and real-world operational data.Projects you may own or contribute to include: enterprise billing & reconciliation engines, accounting workflow automation, AI document processing pipelines, and internal operational tooling.What We’re Looking For (Requirements)We are open to all majors and academic years, and formal software engineering work experience is not required. However, you must have:Hands-on experience building software with Coding Agents: Whether you use Claude Code, Codex, Cursor, or another setup, you must be able to clearly articulate your workflow—how you use AI to build, debug, and verify your code.Solid engineering fundamentals: Proficiency in at least one programming language, with the ability to read, reason about, and modify existing codebases.Core web concepts: A working understanding of Git, APIs, relational databases, and web application architecture.High agency in ambiguity: The ability to proactively break down ill-defined problems, research solutions, and test hypotheses without waiting to be spoon-fed complete specs.Accountability & clear communication: Ownership over your delivery quality, and the habit of communicating progress, risks, and blockers early and clearly.Nice-to-Haves (Bonus Points)Experience building and deploying a live product from scratch. (show us your Github!)Familiarity with TypeScript, Postgres, GCP.Prior exposure to B2B SaaS, workflow automation, or LLM applications.Open-source contributions, side projects, hackathon wins, or experience iterating on a product with real users.Strong curiosity about enterprise software, product design, and early-stage startups.What We ValueWe don’t evaluate candidates by LeetCode scores or how many framework APIs you have memorized. Instead, we look for:Ownership: You push problems forward proactively and take pride in the final outcome.Engineering Judgment: You know what can be delegated to AI—and what strictly requires human verification and architectural rigor.Product Sense: You care deeply about solving the user's actual underlying problem, not just closing tickets.Learning Velocity: You pick up unfamiliar tools and domain knowledge rapidly and apply them immediately.Trust: You uphold high standards for code quality, data security, confidentiality, and transparent communication.Compensation & StructureCompensation: NT$280 – NT$400 / hour (commensurate with technical capability and track record of shipping).Commitment: at least 16 hours/week (flexible scheduling coordinated around your classes/commitments).Work Mode: Taipei-based team collaboration (hybrid mix of in-person sessions and remote work).Structure: An initial 8-12 week onboarding and evaluation cycle, with a structured midpoint review at Week 4.Growth Path: High performers will have the opportunity to extend the internship or convert into a core full-time role as we scale.How to ApplyPlease submit your resume and one software project you've built or substantially contributed to.GitHub repositories, live demos, or technical write-ups are welcome. Private projects are also fine—just describe what you built and your specific contributions.In a few sentences, tell us:What did you build, and whose problem did it solve?How did you use AI Coding Agents during development?No lengthy cover letter required.We'll explore your engineering decisions, debugging process, and approach to validating AI-generated code during the interview.We're looking for builders who use AI to turn real-world problems into working software.