Senior Full Stack Engineer

This role is remote from within South Korea. Applicants must already hold a visa permitting this work.

About the Role

As a Senior Full Stack Developer at Stitch.ai, you will play a critical role in building and launching an AI-native workspace powered by Stitchy, our resident AI teammate.

You will design, develop, and optimize the user-facing applications and backend services that support Stitch.ai’s native email, calendar, tasks, notes, contacts, files, knowledge, and automation capabilities. You will also help build the agentic infrastructure that allows Stitchy to understand context, select and use tools, coordinate multi-step work, maintain working memory, respond to events, and take reliable action across the platform.

This role sits at the intersection of conventional full-stack engineering and production AI systems. It requires strong experience with Angular, TypeScript, Node.js, NestJS, Express.js, and PostgreSQL, together with hands-on experience building agent harnesses, orchestration systems, tool-calling agents, or other LLM-powered applications.

You will work on problems including agent orchestration, context management, memory and retrieval, event-driven execution, tool selection, failure recovery, evaluation, observability, security, latency, and AI inference cost.

Key Responsibilities

Full-Stack Product Development: Design, develop, test, and deploy user-facing features across our Angular frontend and Node.js, NestJS, and Express.js backend services.

Agent Harness Development: Build and improve the software surrounding our AI models, including context assembly, prompt and instruction management, tool availability, execution controls, state management, model routing, structured outputs, and response handling.

Agent Orchestration: Design systems that allow Stitchy to plan and perform multi-step work across native applications while maintaining clear execution state, dependencies, permissions, and completion criteria.

Tool Use and Action Execution: Develop reliable mechanisms through which agents discover, select, call, and coordinate application tools and APIs. Ensure tool calls are validated, observable, permission-aware, and safe to retry.

Memory and Context Management: Build systems that give Stitchy the right information at the right time without overwhelming the model context. Work with conversation memory, application data, retrieval, embeddings, metadata, summaries, and context-compaction strategies.

Agent Reliability and Safety: Implement practical controls for duplicate actions, automation loops, stale context, partial execution, timeouts, retries, conflicting state, and other failure modes unique to agentic systems.

Evaluation and Observability: Create testing and evaluation systems for non-deterministic AI behavior. Instrument agent decisions, tool calls, token consumption, latency, errors, and task outcomes so behavior can be understood and continuously improved.

Model Integration and Optimization: Integrate and evaluate large language models from multiple providers. Improve model selection, inference cost, latency, caching, and context efficiency while preserving response and task quality.

Backend and Microservices Architecture: Build and maintain modular services capable of supporting complex data flows, asynchronous processing, background jobs, real-time updates, and high service reliability.

Database and Data Architecture: Design, optimize, and maintain PostgreSQL data models for a multi-application workspace with interconnected user data, agent state, application metadata, and knowledge relationships.

API and Event Development: Develop secure, well-documented REST APIs and event-driven services that support communication between applications, backend services, and AI orchestration components.

Performance and Scalability: Monitor and improve application performance, database efficiency, queue processing, search performance, and system resilience as Stitch.ai grows.

Security and Privacy: Apply strong security practices to authentication, authorization, tenant isolation, encryption, sensitive-data handling, agent permissions, and auditable AI actions.

Code Review and Mentorship: Conduct thoughtful code reviews, establish engineering standards, and mentor other developers in application architecture, agentic systems, testing, and maintainable software design.

Cross-Functional Collaboration: Work closely with Product, UX/UI, DevOps, and AI engineering to translate product behavior into dependable technical systems. Clearly document architecture, interfaces, decisions, and operational requirements.

Required Skills and Qualifications

Professional Experience: At least seven years of professional software-development experience, including substantial ownership of production web applications and backend systems.

Full-Stack Expertise: Strong professional experience with TypeScript, Angular, Node.js, NestJS, Express.js, and PostgreSQL.

Production Agentic Development: Hands-on experience building or operating production-grade LLM applications, AI agents, agent harnesses, or multi-step tool-using systems.

Agent Orchestration: Experience coordinating model calls, tool execution, state transitions, multi-step tasks, asynchronous events, retries, and failure recovery.

LLM Tool Use: Experience with function calling, structured outputs, tool schemas, execution validation, and integrating agents with internal or external APIs.

Context and Memory Systems: Understanding of context-window management, retrieval, summarization, embeddings, vector search, working memory, and persistent agent state.

AI Reliability: Experience addressing non-deterministic behavior through evaluations, tracing, guardrails, testing, observability, idempotency, and controlled execution.

Backend Architecture: Strong understanding of microservices, asynchronous processing, event-driven systems, queues, caching, and complex application data flows.

Database Engineering: Strong PostgreSQL knowledge, including schema design, indexing, query optimization, transactions, concurrency, and data integrity.

API Engineering: Experience designing secure and maintainable REST APIs, service contracts, authentication systems, and integrations.

Production Operations: Experience diagnosing performance, reliability, and scalability problems across frontend, backend, database, queue, and infrastructure layers.

Engineering Judgment: The ability to turn complex product behavior into practical, maintainable systems without unnecessary architectural complexity.

Communication: Strong written and verbal communication skills, including the ability to explain technical decisions and collaborate closely with product and design teams.

Preferred Skills

• Experience with Model Context Protocol (MCP), agent tool registries, tool discovery, or dynamic tool selection.
• Experience with Anthropic Claude, Google Gemini, OpenAI models, or other major LLM APIs.
• Experience designing agent evaluation suites, simulations, regression tests, tracing, or production quality metrics.
• Experience with prompt versioning, model routing, fallback strategies, semantic caching, and token-cost optimization.
• Experience with retrieval-augmented generation, embeddings, pgvector, semantic search, and hybrid retrieval.
• Experience with RabbitMQ, Kafka, or similar message brokers.
• Experience with Redis or other caching and distributed-state systems.
• Experience with Elasticsearch or OpenSearch.
• Experience with Docker, Kubernetes, CI/CD pipelines, and automated testing.
• Experience with Akamai Cloud/Linode, AWS, Google Cloud, OVHcloud, or comparable infrastructure platforms.
• Knowledge of OAuth, JWT, encryption, role-based permissions, data isolation, and security practices for multi-tenant applications.
• Experience using AI coding agents and developing effective engineering workflows, harnesses, specifications, and review processes around agent-generated code.
• Experience working in an early-stage startup and taking features from product definition through production deployment.

What We Are Looking For

We are looking for an engineer who is equally comfortable building a polished Angular interface, designing a PostgreSQL data model, debugging an asynchronous backend service, and reasoning about why an AI agent selected the wrong tool or repeated an action.

The right person will bring strong engineering fundamentals, practical agentic-development experience, product judgment, and a desire to help build a genuinely new kind of software as Stitch.ai moves from stealth toward launch.