All Jobs

Detecta Language Agency is currently recruiting native Japanese speakers for an English conversational recording project.

The purpose of this project is to collect 100 hours of high-quality English conversational audio spoken with a natural Japanese accent. This data will support the development of AI models for accent recognition and multilingual speech technologies.

Participants will engage in natural, unscripted English conversations while maintaining their authentic Japanese accent. Conversations should reflect real-life communication, including natural flow, interruptions, and overlaps.

Project Type
English conversational recording with Japanese-accented speech (non-native English).

Job Type

• Part-time
• Fully remote
• Flexible schedule
• Task-based completion

Project Requirements

• Native Japanese speakers (Japan-based)
• Able to speak English fluently (Japanese accent acceptable and expected)
• Total participants: 100
• Age range: 18–60 years
• Balanced gender ratio (Male/Female ~1:1)
• Natural speaking ability (no scripted or read speech)

Recording Requirements

• Conversations must be between 2 participants (pair-based)
• Dual-channel recording (separate tracks for each speaker)
• Natural, free-flowing English conversations
• Interruptions and overlaps are allowed and encouraged
• Each segment duration: 5 to 30 minutes
• Total recording per group: maximum 2 hours

Technical Requirements

• Recording device: Mobile phone or professional setup
• Audio format: Mobile: 32kHz, 16-bit, mono
• Professional setup: 48kHz, 16-bit, mono
• Multi-channel recording preferred
• Quiet indoor environment (noise < 50dB, no echo)
• No background noise (traffic, music, typing, etc.)

Content Guidelines

• Conversations must be fully in English
• Natural speaking style (no reading or acting)
• Avoid sensitive topics (violence, political, explicit content)
• Both participants must actively contribute
• Clear speech required (no heavy distortion or unclear pronunciation)

Quality Requirements

• Word accuracy target: 98%
• No clipping, distortion, or echo
• Clean and consistent audio quality
• Invalid or low-quality recordings will be rejected

Payment & Process

• Payment details will be shared at project onboarding
• Selected candidates will be contacted by our team
• Rework option available for rejected recordings
• Voice authorization required before participation