About the Role
We are hiring Language Experts / Lexical Transcribers to support a voice bot annotation project involving real customer speech. The role requires accurate transcription and audio segmentation in accordance with detailed client annotation guidelines. The project will begin with a calibration team and progressively scale across additional Indian languages.
Languages : Hindi, Tamil, Telugu, Malayalam, and Kannada.
Qualification :Finance graduates with prior transcription experience and a typing speed of more than 50 words per minute.
Location : Pune , Maharashtra.
Mode of Work : Onsite.
Key Responsibilities
- Transcribe customer-side voice bot call recordings accurately in the assigned target language.
- Segment only the primary customer speech and follow the prescribed segmentation rules, including minimum segment duration and non-overlapping segments.
- Produce lexical transcription without punctuation or special symbols and use the appropriate target-language script.
- Write numbers, dates, digits, abbreviations, and English loan words in the required lexical format while maintaining consistent spelling.
- Apply contextual correction only when the intended word is unambiguous from the sentence context; otherwise preserve what is actually spoken or use the prescribed unintelligibility handling.
- Identify and appropriately handle distorted, unintelligible, IVR, non-target-language, background-speech, and overlapping-audio cases as defined in the guidelines.
- Use the tag only for genuinely unintelligible content and follow project quality limits.
- Incorporate calibration and QA feedback promptly and maintain agreed quality and turnaround expectations.
- Escalate unclear or exceptional cases through the defined operations process rather than making unsupported assumptions.
Required Qualifications
- Graduate in Finance
- High proficiency in at least one of the required project languages, with strong command of the language's standard written form and script.
- Prior transcription experience is required; experience with audio annotation, speech data, voice bot data, or linguistic annotation is preferred.
- Typing speed of above 50 words per minute (WPM).
- Strong listening ability and attention to detail, especially with real-world call audio containing noise, accents, pauses, or audio artifacts.
- Ability to consistently follow detailed annotation guidelines and apply the same spelling and formatting conventions across batches.
- Comfort working against productivity, quality, and turnaround-time expectations in a structured production environment.
Preferred Skills
- Experience with lexical transcription, speech-to-text validation, linguistic data labeling, or conversational AI datasets.
- Familiarity with financial-services terminology or customer-service call recordings.
- Experience identifying audio quality issues, overlapping speech, IVR leakage, and non-target-language segments.
- Strong ability to distinguish intelligible speech from genuinely unintelligible content.
What Success Looks Like
- High-quality, consistent transcripts that follow the client's lexical transcription guidelines.
- Accurate customer-only segmentation with minimal avoidable clipping or inclusion of non-target speech.
- Timely completion of assigned volumes and prompt incorporation of review feedback during calibration and production.
- Careful use of exceptions such as , dropped/rejected audio, and non-target-language handling.
- Reviewing guidelines, Query resolutions, Consulting references while actively working on tasks.
