Offer summary

Qualifications:

Engineering experience in systems or backend work, particularly in API design and orchestration., Proficiency in Node.js, Typescript, NextJS, and modern service-oriented architecture., Ability to architect systems with thoughtful abstractions and iteration in mind., Strong communication skills and a collaborative mindset..

Key responsibilities:

Build APIs and orchestration layers connecting AI reasoning to developer workflows.

Design and implement the MCP API for structured, context-aware data delivery.

Collaborate on the evolution of the Composition Advisor for validation and prototyping.

Document work clearly and participate in decision-making and prioritization.

Job description

Join Knapsack to help build intelligent infrastructure that connects design, code, and AI - and redefines how digital products are made.

About Knapsack

At Knapsack we’re building an AI Product Engine for digital product creation. Our mission is to eliminate the gap between what teams design and what they build. We do that by grounding every tool, agent, and workflow in a living, intelligent, source of truth.

We’re not building another AI wrapper or a glorified assistant. We’re building a foundational platform - MCP (Model Context Protocol) server - that powers the next generation of intelligent product development. Our Composition Advisor will help teams validate, compose, and evolve UI intelligently. Our Ingestion Agent pulls meaningful patterns from codebases, Figma, and other artifacts to bootstrap structured systems from chaos.

We’re not solving a well-understood problem. We’re inventing new territory.

Overview

We’re hiring a Software Engineer - Intelligent Product Systems who’ll help us bring that vision to life. This is a foundational role, early in this evolution, we’ll be honest - we don’t have all the answers yet. We’re experimenting. We’re figuring things out. But we know that getting the infrastructure, context, and delivery pipeline right is the key.

You’ll be a key contributor in shaping the technical foundations of how AI-powered product development actually works - and works well.

What we think you’ll do

Depending on your strengths, experience, and what we discover together, you might:

Build APIs and orchestration layers that connect AI reasoning to developer workflows (CLIs, IDEs, Git hooks, etc)
Design and implement our MCP API, providing structured, context-aware data (tokens, components, rules, and guidance) to internal and external agents
Help build our first ingestion agent, capable of extracting meaningful system data from Figma, Storybook, and product codebases
Collaboration on the evolution of the Composition Advisor - a validator and prototyping assistant powered by structured context, not hardcoded rules
Shape the early infrastructure for prompt delivery, inference routing, and feedback loops
Own and evolve the data schema that serves as the backbone for all MCP-based interactions
Document your work clearly and share in decision-making and prioritization. This is a buildout, not an spec handed down from above

Some of the things we think you’ll need

We need someone who’s flexible, pragmatic, and comfortable in uncharted territory. We think these kind of experiences will help:

Engineering experience across systems or backend work (API design, orchestration, data pipelines, etc)
Fluency with Node.js, Typescript, NextJS, and modern service-oriented architecture
Ability to architect systems with thoughtful abstractions and iteration in mind
Familiarity with developer tools and structured design data (tokens, components, system schemes)
Curiosity about LLMs and how context, prompts, and inference shape behavior
Comfort working independently, defining your own structure, and asking smart questions when direction is fuzzy
Strong communication skills and a collaborative, low-ego mindset

Bonus points for

Experience with LLM tooling: embeddings, RAG pipelines, prompt chaining, or inference orchestration
Familiarity with Figma APIs, Storybook, or other design artifact ingestion workflows
Infastructure experience: observability, serverless deployments, background jobs, or plugin systems
Contributions to AI tools, GitHub bots, or other developer productivity enhancers

Why it matters

This is a systems role at the edge of AI, design systems, and developer tools. The work you’ll do doesn’t exist yet. And while that’s exciting, it’s also hard. We can’t promise perfect specs or polished requirements. We’re building the plane while flying it - thoughtfully, humbly, and with a real chance to shape how teams will build software in the next decade.

If you’ve ever said;

“I want to work on something nobody has solved before”

“I like building useful infrastructure, not flashy demos”

“I want my work to outlast the trend cycle”

…then we should talk.