Tech Lead Engineer (Voice/Speech)

About JAPAN AI

JAPAN AI Inc. is an AI company founded in April 2023 as part of the GENIEE Group, whose parent company, GENIEE, Inc., is publicly listed.

Our ambition goes far beyond providing AI chatbots. We are building the next generation of core enterprise systems: an “enterprise brain” that connects all of a company’s SaaS applications and enables AI to carry out business tasks autonomously. Our core product, JAPAN AI STUDIO, has been adopted by a total of 400 companies and more than 19,000 users, and our revenue has grown approximately fivefold year over year.

Learn More

• Company website: https://japan-ai.co.jp/

• Company overview deck: https://www.docswell.com/s/geniee/Z828R2-japan-ai_company-deck

• Engineering blog: https://engineering.japan-ai.co.jp/

• Careers: https://japan-ai.co.jp/recruit/

Why We’re Hiring

JAPAN AI SPEECH transcribes meetings and sales conversations in real time, distinguishes between speakers, and organizes decisions and action items into structured outputs. In 2026, we are rolling out a new speech recognition engine and cross-meeting analytics, making this one of our fastest-growing product areas.

But the technical challenges now go well beyond what the label “meeting transcription tool” suggests. We need to balance three competing priorities—real-time performance, recognition accuracy, and GPU inference costs—while supporting customer-specific terminology, translation into more than 60 languages, and requirements to process data exclusively in Japan-based cloud regions. We currently lack a technical leader who can take ownership of these challenges from the architecture stage and guide the team through them. We are hiring for this role to establish a dedicated technical decision-maker for Speech.

What You’ll Do as Tech Lead

Own the technical strategy and architecture of our speech processing platform

Design the end-to-end processing pipeline, from streaming audio ingestion through speech recognition, speaker diarization, and summarization. You will own the decisions about how to balance accuracy, latency, and inference costs: what to prioritize, to what extent, and how to achieve it. This also includes technology selection, technical debt reduction planning, and designing for scalability and cost efficiency.

Lead the design and implementation of product features

Lead the hands-on design and development of cross-meeting analytics, LLM-powered summarization and key discussion point extraction, audio ingestion from Zoom, Teams, and Meet, and integrations with business tools such as Salesforce and Slack. You will work closely with the Product Manager to refine requirements and incorporate enterprise requirements into the architecture, including compliance with ISMS and PrivacyMark (P Mark) requirements, data encryption, and processing within Japan-based cloud regions.

Example scenario: Speaker diarization accuracy has plateaued for meetings with many participants. You redesign the pipeline to combine an audio waveform-based separation model with downstream LLM-based corrections, and introduce batched GPU inference to improve accuracy while reducing inference costs.

Provide technical leadership for the team

Raise engineering quality through design and code reviews, mentor team members, and improve development processes. You will also help build the team by participating in technical interviews.

The Team

Our engineering organization has approximately 60 members, within a company of 256 employees. Non-Japanese nationals make up 44% of the engineering team.

The Speech team has five members and works closely with the Product Manager, the AI Platform team responsible for our GPU and Kubernetes infrastructure, and Product Engineers.

Required Qualifications

• A bachelor’s degree in computer science, software engineering, artificial intelligence, machine learning, mathematics, physics, or a related field, or equivalent professional experience.

• At least five years of professional experience as a software engineer.

• At least three years of experience as a technical lead or architect, including responsibility for technology selection, architectural decisions, and technical guidance for team members.

• Experience designing and operating large-scale backend systems with high availability, scalability, and cost efficiency in mind.

• Hands-on experience with cloud platforms such as GCP, AWS, or Azure, and container technologies such as Docker and Kubernetes.

• Professional experience in at least one of the following:

  • Real-time or streaming processing using technologies such as WebSocket, gRPC, Kafka, or Pub/Sub.

  • Designing and operating machine learning inference infrastructure in production.

• Experience working at a startup, a growth-stage company, or in a comparable environment.

• Proficiency in at least one of the following languages:

  • Japanese: Business-level proficiency sufficient to interview and work in Japanese.

  • English: Fluent.

Preferred Qualifications

• Professional or research experience in speech-related fields, such as speech recognition, speaker diarization, or audio signal processing.

• Experience operating and optimizing GPU inference infrastructure, including technologies and techniques such as Triton, TorchServe, ONNX Runtime, batching, and quantization.

• Experience developing products that incorporate LLMs.

• Experience designing and operating MLOps pipelines.

• Experience designing and operating large-scale data platforms, such as BigQuery or data lakes.

• Experience designing and operating systems subject to security requirements such as ISMS or PrivacyMark.

• Experience expanding into audio channels beyond meetings, such as telephony and computer telephony integration (CTI).

• Experience taking a new product from zero to one.

• Full-stack development experience with TypeScript and React.

• Bilingual proficiency in Japanese and English.

Development Environment

• Languages and frameworks: Python for the backend; TypeScript, React, and Next.js for the frontend; Nx.

• AI / LLM: LangChain, LangGraph, JAPAN AI STUDIO SDK.

• Infrastructure: GCP (containers / Kubernetes), Docker.

• Data: BigQuery, PostgreSQL, and a variety of customer data sources.

• Collaboration and development tools: Slack, Confluence, Linear, Google Workspace, GitHub, Notion.

• AI development assistants: Claude Code Max plan, Cursor, ChatGPT, Devin.

• Equipment: Apple Silicon Mac, with support for dual monitors.

Learning, Development, and Wellbeing Benefits

Company-funded AI tools

Access a wide range of AI tools at the company’s expense, including JAPAN AI’s SaaS services, Cursor, ChatGPT, and Claude.

Development tool allowance

We cover the cost of paid development tools you would like to use, up to JPY 30,000 per year.

Book purchase allowance

Purchase technical books and other books for learning at the company’s expense, up to JPY 30,000 per six-month period.

Language learning and certification support

The company covers the cost of Japanese and English language learning programs and professional certifications.

Wellbeing allowance

We cover services and activities that help you recharge, up to JPY 5,000 per month. Eligible expenses include gym memberships, yoga, bodywork treatments, and tickets to aquariums, movies, and theme parks, among others.

Housing allowance

Employees living in eligible areas receive a housing allowance of up to JPY 30,000 per month.

Work Arrangement

Hybrid: three days in the office and two days working remotely per week.

Hiring Process

Application screening → Coding assessment → Interviews (4–5 rounds) → Offer

You will be asked to complete a reference check by the final interview.