AI Backend Engineer (AI Workflow Systems)
π’ Bjak Β· all Bjak jobs
π South Korea
π
Posted 2026-06-29 Β· via Himalayas
π· AI-Backend-Engineer,Backend-AI-Engineer,Senior-AI-Backend-Developer,Mid-Level-AI-Backend-Developer
Apply on original site βBJAK is Southeast Asia's largest digital insurance platform, building AI-powered products that simplify insurance and financial services for millions of users. We use AI and automation to transform real-world insurance workflows such as quotations, policy issuance, endorsements, claims, and customer follow-ups.
We are looking for talented AI Backend Engineers to build the systems that power AI-driven decisioning, automation, and orchestration across our platform. This role sits at the core of our AI stack - between models, backend systems, and real users - where latency, correctness, reliability, and cost directly impact production experience.
This is a fully remote position where you will be part of a global engineering team working across multiple countries to build reliable, scalable AI systems.
Focus
-
Build and operate backend systems that serve AI-powered insurance workflows in production.
-
Design and implement AI orchestration layers that connect models, APIs, workflows, and business logic.
-
Build inference pipelines for LLM-based and AI-assisted automation systems.
-
Optimize latency, throughput, and cost across AI services (caching, batching, streaming, routing).
-
Design stable service boundaries between backend systems, ML components, and product APIs.
-
Implement observability: logging, metrics, tracing, alerting, and incident response workflows.
-
Debug production issues across distributed AI systems and resolve root causes.
-
Collaborate closely with frontend, product, operations, and ML teams to ship end-to-end features.
-
Continuously improve system reliability, scalability, and performance.
Ideal Experience
-
Strong backend engineering experience in production systems.
-
Experience building or operating high-throughput, low-latency services.
-
Familiarity with AI systems (LLMs, embeddings, or AI workflows).
-
Experience with distributed systems and production debugging.
-
Strong understanding of APIs, data flows, and system design principles.
-
Experience with observability tools (logging, monitoring, tracing).
-
Strong ownership mindset and bias toward shipping.
-
Comfortable working in fast-paced, globally distributed teams.
Outcomes
-
AI backend systems run reliably at scale with low latency and high availability.
-
AI workflows are stable, observable, and production-ready across multiple products.
-
System performance improves continuously through real-world feedback and optimization.
-
Production incidents are quickly detected, diagnosed, and resolved.
-
AI capabilities are seamlessly integrated into global customer and internal workflows.
Tech Stack
- Python
- Node.js
-
LLM APIs (OpenAI / Anthropic / open-source models)
- SQL / NoSQL databases
- Kubernetes
- Docker
- Distributed systems tooling
-
Observability stacks (logging, metrics, tracing)
How We Work
We believe strong products are built by small, high-ownership teams.
Engineers at BJAK work closely with product, design, AI, and operations teams to solve meaningful real-world insurance problems. We value technical excellence, speed of execution, and practical decision-making.
Engineers are expected to own systems end-to-endβfrom design and implementation to production reliability and iteration.
Why Join BJAK
-
Build AI-Powered Products - Work on intelligent insurance automation systems.
-
Global Engineering Organization - Collaborate across multiple countries.
-
International Impact - Products used by millions across Southeast Asia and beyond.
-
Learning & Development Budget - Support for continuous growth.
-
High Ownership Culture - End-to-end ownership of engineering systems.
-
Modern Engineering Practices - Focus on scalability and reliability.
-
Career Growth - Fast-moving engineering environment.
-
Competitive Compensation - Attractive salary package.
-
Fully Remote - Work remotely with globally distributed teams.
The Kind of Builder We Want
-
Th