Customer Support & AI Evaluation Specialist

Evaluated and improved over 10,000 AI/LLM-generated customer support responses in English, Hindi, and Hindi-Latin. Reviewed outputs for accuracy, empathy, tone, clarity, relevance, problem resolution, user intent understanding, and instruction following.

Applied client-specific evaluation rubrics to score outputs, identify recurring failure patterns, provide written justifications, and supply corrected responses where needed. Performed factuality, groundedness, hallucination, attribution, context-mismatch, and reasoning-consistency evaluations.

Joined as a Customer Support Agent and was promoted to Customer Support Supervisor. Managed high-volume customer and affiliate queries through live chat, email, and ticketing systems, resolving account, billing, payment, technical, complaint, and escalation cases while emphasizing timely and accurate resolutions.

Supervised a fully remote international support team, provided real-time guidance on difficult cases, reviewed agent quality, coached team members, and monitored CSAT, response time, resolution time, and quality standards. Trained and onboarded agents on product knowledge, SOPs, escalation processes, support tools, and workflows.

Performed multilingual AI content evaluation and linguistic QA for grammar, fluency, semantic accuracy, script compliance, naturalness, and cultural relevance. Worked on prompt evaluation, localization QA, MTPE, translation review, transcription QA, data annotation, and sentiment and toxicity classification projects.

Translated and reviewed millions of words of English-Hindi content for international clients, including e-learning, websites, technical and financial documents, marketing materials, subtitles, films, documentaries, and digital media. Ensured accuracy, terminology consistency, contextual fidelity, fluency, tone, and cultural relevance using client glossaries and style guides.

Completed Hindi and English transcription, speech-data review, and annotation work for voice-recognition and AI initiatives, including Google Home and Microsoft-related projects. Annotated text, audio, image, and language datasets through tasks such as text classification, intent tagging, speaker identification, audio labeling, and data categorization.

Managed multiple freelance projects simultaneously, maintained client communication and confidentiality, and achieved 5-star client ratings and a 100% Upwork Job Success Score.

Delivered customer support through phone, chat, and email for billing, technical issues, account management, and service upgrades. Maintained an average handling time below three minutes while delivering fast, accurate, and empathetic assistance and sustaining strong FCR and CSAT outcomes.

Promoted to Trainer after two years based on performance, product knowledge, and communication skills. Conducted onboarding, live coaching, mock calls, product walkthroughs, and nesting-period support for new agents.

Created training materials, support documentation, and performance evaluation systems. Acted as a liaison between frontline staff and operations to improve workflows, reduce escalations, and maintain consistent service quality through CRM tools, ticketing systems, and SOPs.