MD TABISH

Design and test prompts for instruction following, accuracy, relevance, and response consistency. Document failure patterns, edge cases, and quality issues; support bilingual and guideline-driven evaluation.

Evaluated Large Language Models through structured testing and rubric-based assessment of reasoning, factuality, instruction following, and safety. Performed prompt optimization, comparative response evaluation, and conversational AI evaluation across multiple AI projects.

Complete task-based AI work in Mercor's task workspace under detailed project guidelines and quality standards. Evaluated and rated multi-turn AI model responses across five prompt categories against structured rubric criteria.

Engineered Python automation scripts for SSH, Telnet, and RDP to automate bulk password resets for 1000+ nodes, reducing manual effort by 97%. Integrated ServiceNow REST APIs to auto-process user access tickets, reducing resolution time by 40%.

Performed multimodal annotation across text, image, OCR, video, and audio for AI training and evaluation workflows. Delivered Search Quality Rating, Urdu OCR/ASR review, and auto-dubbing evaluation at approximately 99% accuracy.

Contributed to language and content evaluation, including Hindi in Latin script, proofreading, generated-content review, multilingual annotation, and language validation against project quality standards.

Evaluated AI-generated language outputs for linguistic accuracy, naturalness, contextual relevance, and cultural appropriateness using Hindi, English, and Urdu expertise. Documented review decisions and provided structured feedback for multilingual model improvement.

Contributed to text-based AI alignment tasks using project rubrics and documented evaluation criteria. Applied consistent judgments and recorded concise justifications to support training-data quality.

Compared model responses across conversational attributes, rated outputs independently, and wrote rationales and direct preference comparisons. Completed calibration and incorporated submission feedback into guideline-driven tasks.

Designed and optimized prompts for LLMs to improve response quality, instruction adherence, and conversational consistency. Evaluated AI-generated outputs using structured rubrics to identify hallucinations, reasoning gaps, and factual inaccuracies.

Supported AI model training, AI content operations, and LLM evaluation workflows for an enterprise AI platform. Applied structured analytical thinking, content-quality review, and process-oriented improvements.

Taught music through structured classroom and practical lessons, adapting instruction to learner skill levels. Guided students in music fundamentals, singing, and rhythm with individualized attention.

Led React Native mobile application development projects, handling project architecture, sprint planning, and technical decisions. Managed projects independently, coordinated with clients, performed code reviews, and mentored team members.

Developed and automated ETL pipelines using Python and SQL, improving data processing speed by 50%. Analyzed large-scale telecom datasets and built interactive dashboards in Tableau and Power BI.

Provided structured, rubric-style teaching for school-level students. Built communication, mentoring, and feedback skills transferable to AI evaluation and human-feedback work.

Led operational teams against productivity, quality, and SLA benchmarks. Supported daily coordination, performance monitoring, and process improvement.

Developed mobile applications for HRMS, ERP, and Retail solutions. Integrated APIs and real-time synchronization features, improved application performance and UI responsiveness.

Handled international customer interactions, communicated solutions clearly, and maintained service-process standards.

Managed high-volume customer-service operations, maintaining 10,000+ records at approximately 99.8% accuracy. Handled issue resolution, documentation, and regulated service and quality standards.

Built AI-powered systems using OpenAI APIs, LLM pipelines, Prompt Engineering, LangChain, and RAG architectures. Designed scalable backend workflows integrating vector databases, REST APIs, Python, and real-time AI systems.

Developed AI-powered applications using Python, LLMs, REST APIs, and modern AI frameworks. Built intelligent automation workflows integrating LLM APIs, prompt engineering, and backend services.

Evaluated AI-generated responses for accuracy, relevance, reasoning, code quality, and instruction following. Reviewed technical tasks involving software engineering, Python, JavaScript, APIs, Git, Docker, and AI workflows.

Authored software-engineering tasks with objectives, metadata, environment summaries, difficulty/effort estimates, oracle and verification strategies, binary success conditions, partial-score strategy, anticipated exploits and resource/network requirements.

Label and quality-review video-based AI training datasets using structured annotation guidelines and rubric-based evaluation. Review visual and contextual content for accuracy, consistency, and alignment with project-specific quality standards.

Contributed to content-language labeling, search side-by-side review, web/mobile/desktop scenario testing, map-data checks, data validation, and AI content evaluation.

iMerit: household-activity video data collection. TELUS International: photo and video annotation. Upwork: remote AI/data annotation, content evaluation assignments under client-specific quality standards.

Currently pursuing a Master of Business Administration, expected to complete in 2027.

Grade: 7.5 CGPA

Pursued a Bachelor of Technology in Computer Science and Engineering, achieving a CGPA of 7.5.

Completed a Master of Arts in Political Science from 2022 to 2024.

Completed a Bachelor's in Information Technology from March 2017 to March 2021.

Earned a Bachelor of Arts in Political Science from 2018 to 2021.

Earned a Bachelor of Computer Applications in Computer Science from 2013 to 2016.