Job Description

Career Opportunity: Agent Harness Engineer in Japan | Tokyo (Hybrid) !

■ Agent Harness Engineer (English Required, Japanese is a plus)

■ Company Overview

A fast-growing, engineering-first technology company with over 80% engineers, focused on driving innovation through strong technical leadership and continuous learning.

With a diverse global team across 25+ countries, the company works in small, agile teams delivering high-impact solutions.

Key domains include:

FinTech: Modernizing financial systems, payment platforms, and core banking

PharmaTech: Building SaaS solutions to enhance R&D productivity

Design and build the core execution infrastructure (“Agent Harness”) that enables AI agents to operate securely, efficiently, and at scale.

You will develop systems powering hundreds of workflows on an in-house AI platform used by enterprises

■ Your Role and Responsibilities 

The Agent Harness is the execution & control layer for AI agents—handling orchestration, memory, guardrails, and model routing—going beyond traditional backend engineering into a new AI-native domain. 

Agent Harness Development

●Design & build agent execution engines (state machines / graph runtime)

●Develop Agent SDKs for internal engineers

●Implement session management, checkpoints, recovery systems

●Build guardrails & policy execution systems

AI/ML Integration 

●Model routing across multiple LLM providers

●Design RAG, memory systems (long-term & working memory)

●Optimize inference (latency, cost, caching)

●Collaborate with research teams to productionize new AI methods

Orchestration & Performance

●Workflow orchestration, queuing systems

●Autoscaling, load balancing, cost optimization

Reliability & Operations

●Maintain 99.9%+ uptime

●Incident handling & system improvements

●Access control & data security design

 

 

■ Experience and Qualifications

●Bachelor’s degree (or equivalent) in CS, AI, Engineering, or related field

●5+ years of backend engineering experience

●Strong experience in Python (production systems)

●Experience building LLM/AI-driven systems

●Experience with distributed systems & APIs (REST/gRPC)

●Language:English (Business level) OR Japanese (Fluent)

 

Tech Stack:

Backend: Python, Go

Frontend: TypeScript, React, Next.js

Infra: GCP, Kubernetes, Docker, Terraform

Messaging: Kafka, Pub/Sub

Monitoring: Prometheus, Grafana, OpenTelemetry

Tools: GitHub, Slack, Notion, Confluence

 

■ Good Reasons to Join

●Work at the intersection of AI × backend engineering

●Build core infrastructure powering enterprise AI systems

●High ownership in architecture & technical decisions

●Collaborate in a fast-growing, global AI environment

 

■ Work Location

Tokyo, Japan (Hybrid)

👉 Interested or know someone suitable? Let’s connect!

📩 Share your updated CV at kanika.pal@talisman-corporation.com

📅 Limited interview slots available – early applicants preferred   

Details will be provided during the meeting.