Native Bengali.
Zero Translation Drift.

Services

• Voice Training — Clear, neutral Bengali pronunciation for voice model training
• Text Annotation — Bilingual (Bengali ↔ English) content review and categorization
• Cultural Context Annotation — Regional variations, idioms, cultural nuances
• LLM Evaluation — Model output quality assessment for Bengali language models
• Localization — Software, apps, and content adapted for Bengali-speaking users
• Data QA — Structured, consistent, machine-readable annotation outputs

What Makes Me Different

• Native Bengali — No translation layer, no cultural drift. Fluent across Bangladesh and West Bengal regions.
• Fluent English — Clear async communication with global teams. No miscommunication on task requirements.
• Technical Background — BSc CSE, full-stack dev. I understand data formats (JSON, CSV, XML), APIs, and QA processes.
• Reliable — Quiet workspace, quality microphone, reliable internet. Available 20-30 hrs/week, GMT+6.
• AI-Native — I use LLMs daily. I understand model training, prompt engineering, and evaluation. Not annotating blindly.

Availability

• Hours: 20-30 hours per week
• Timezone: UTC+6 (Dhaka, Bangladesh)
• Overlap: US morning (8AM-12PM ET) + EU afternoon (2PM-6PM CET)
• Start date: Immediate
• Equipment: Quiet workspace, quality microphone, reliable internet

Start a Project

Tell me about your Bengali AI needs. I'll send a detailed proposal with timeline and cost estimate within 24 hours.