Full-Time Backend Engineer

We are a international innovative Korean travel-based search engine startup revolutionizing how users discover, plan, and book travel experiences across South Korea and beyond. Our platform powers personalized itineraries, real-time recommendations, and seamless mobile experiences for global travelers.

(<https://konnect.kr/about>) (<https://koreatechdesk.com/k-startup-grand-challenge-ksgc-comeup-2025-winner-indias-konnect>)

As a Senior Backend Engineer, you will be responsible for the core architecture that powers our next-generation travel discovery engine. You won't just be CRUD-ing data; you will be building high-performance systems that understand traveler intent. You will bridge the gap between traditional databases and modern LLM-driven architectures to provide hyper-personalized itineraries, real-time availability, and "human-like" travel assistance.

This is a full-time onsite (with WFH options) role based in Pangyo, South Korea, with opportunities to shape the future of travel tech. Please note that our communication language shall be English. Please note that we will offer competitive salary based on experience.

-Architect Hybrid Search Systems: Design and maintain systems using Reverse Indexes (Elasticsearch/Solr) for exact matching (e.g., "Hotel in Jeju") and Vector Databases (Milvus/Pinecone/Weaviate) for semantic discovery (e.g., "Cozy places near the beach for a solo writer").

-Build RAG Pipelines: Develop and optimize Retrieval-Augmented Generation (RAG) workflows to ground our AI agents in real-time travel data, ensuring recommendations are factually accurate and up-to-date.

-Graph-Based Personalization: Leverage Graph Databases (Neo4j/ArangoDB) to map complex relationships between users, destinations, local attractions, and transit hubs.

-Optimize Embeddings: Implement and fine-tune Embedding Algorithms to transform unstructured travel content (reviews, blogs, images) into searchable high-dimensional vectors.

-Engineering for Performance: Tackle the "Context Switching" challenge within LLM calls—minimizing latency and managing token windows to ensure a fluid user experience even during complex multi-step travel planning.

-Scalable API Design: Build robust, concurrent backend services (Go, Java, or Python) capable of handling the high-traffic surges typical of Korean holiday seasons (Chuseok, Seollal).

-Core Backend: 5+ years of experience with high-concurrency languages (Go, Java, or Node.js).

Search & Retrieval: Deep understanding of Inverted/Reverse Indexes and how to tune BM25 or Lucene-based search.

-Vector & Graph: Hands-on experience with Vector Databases and Graph Theory (specifically for recommendation engines or knowledge graphs).

-AI/ML Integration: Proven track record of deploying RAG systems and working with LLM orchestration frameworks like LangChain or LlamaIndex.

-Data Science Lite: Proficiency in Embedding Algos (Sentence-Transformers, OpenAI embeddings) and understanding how to evaluate retrieval quality (NDCG, Recall@K).

-Infrastructure: Experience with Cloud-native environments (AWS/GCP), Kubernetes, and distributed caching (Redis).

-Onsite work in Pangyo with optional work-from-home flexibility

-English-based global working environment

-Competitive salary based on experience

-Flexible and open vacation policy

-Support for international conferences and technical training

-Learning budget for books and courses

-Latest hardware and development tools provided

-Meal allowance or provided meals

-Full compliance with Korean social insurance (National Pension, Health Insurance, Employment Insurance, Industrial Accident Insurance)

-Family and special occasion support