Senior Full Stack Engineer, GenAI & Web Tools
Overview Corporation
- Taipei, Taiwan
- Hybrid
- Senior Level
- Posted
is transforming manufacturing quality control with an advanced edge-enabled AI vision system that combines cutting-edge deep learning with ease of use and quick setup. Our system handles a wide array of inspection tasks—from subtle assembly verification to defect detection and orientation checks—delivering real-time, high-accuracy inspections that seamlessly integrate into existing production lines.
Trusted by leading manufacturers like Ford, Honda, Toyota, SpaceX, Milliken, and Flex-N-Gate, our cameras enable faster throughput, reduced scrap, and lower inspection costs—without the complexity and expense of traditional vision systems.
to build the web tools and GenAI applications that turn our edge-AI vision platform into products people use every day—both the customer-facing tool suite and the internal tooling that lets our teams move faster.
This is a hands-on, full-stack role. You'll own tools end-to-end, from a React/TypeScript interface down to the Python services behind it, and you'll be building on the frontier: LLM-powered assistants, synthetic data generation, and agent-style workflow builders that sit on top of our vision system. Think products like Auto-Defect Creator Studio, and
You'll operate with high autonomy in a small, fast team: you pick the approach, ship quickly, and iterate with real users.
About Overview Corporation
- You come in as a senior builder, not a ticket-taker. You define the tools, pick the stack, and decide how they get built. We hire smart people and get out of their way.
- GenAI is core to what we're building, not a bolt-on. You'll ship real LLM-powered products to real manufacturers—and see them used on the factory floor.
Responsibilities
- Design and build LLM-powered tools—AI copilots and expert assistants (documentation search, guided consultation), synthetic training-data generation, and agent-style flow builders—that make our vision platform smarter and easier to use.
- Own web tools end-to-end, from React/TypeScript frontends to Python backends and APIs, spanning both customer-facing products and internal utilities (e.g., optics calculators, recipe analysis, log/telemetry tooling).
0→1 Product Building
- Rapidly prototype new tools, get them in front of users, and iterate. You'll turn rough ideas into shipped products, then sharpen them based on real usage.
- Connect the tools to our edge-AI stack—camera fleet, inference outputs, recipes, and firmware—so they work against real inspection data, not toy examples.
- Sweat performance, reliability, and UX. These tools are how customers and colleagues experience Overview; they should feel fast, clear, and trustworthy.
Requirements
- 4+ years building full-stack web applications, with a proven track record of shipping products end-to-end at a Senior or Staff level.
- Strong proficiency in
- for building modern, polished web interfaces.
- Strong proficiency in
Python
- A product mindset: you care about the user, make good judgment calls under ambiguity, and can drive a tool from idea to shipped without a detailed spec.
- Professional experience with (or a strong passion for) using AI agent tools to architect solutions and solve difficult problems efficiently.
Preferred
- Device Management Platform
- for backend services, APIs, and data/tooling work.
- Experience building
LLM-powered / GenAI product features
- —calling LLM APIs, prompt engineering, RAG, or agent workflows—or a strong passion and demonstrated aptitude for it.
- Hands-on experience with RAG infrastructure, vector databases, or agent-orchestration frameworks.
- Experience with node-based / flow-builder UIs (e.g., React Flow), data visualization, or design systems.
- Familiarity with computer vision, manufacturing, or industrial automation domains.
- Docker and cloud deployment experience.
Benefits
- —plus the utilities that make the platform easier to run.
- Join a fast-growing, fast-moving, and profitable startup.
Work Smarter, Not Harder
We don't just build AI; we use it. If you can lean on an AI agent to automate the grunt work, we'll high-five you, not ask why you aren't typing. We measure leverage, not hours.
jobapplier