Farhan Ishrak Rafi

Farhan Ishrak Rafi

About

Detail

Prompt Engineer | AI Trainer | STEM Expert | M.Sc. Student, Biochemistry & Biology, RU | Research Assistant, BRB | Former Vice President, ASM-RU
Rajshahi, Rajshahi Division, Bangladesh
Contact Farhan regarding: 
work
Full-time jobs

Timeline

work
Job
school
Education

Résumé

Jobs verified_user 0% verified
  • P
    Biology & Chemistry Expert (AI Trainer)
    Poindexter Labs
    Oct 2025 - Current (1 year)
    I contributed as a biology subject matter expert on frontier AI evaluation projects, authoring expert-level benchmark prompts to test advanced reasoning, factual precision, and visual interpretation in large language models. Designed original, machine-verifiable tasks across molecular biology, biochemistry, microbiology, genetics, neuroscience, anatomy and physiology, and research methods, aligned with HLE, Omniscience, and critPT benchmark styles. ●●Key Responsibilities:●● ●Authored original biology prompts (text-only and multimodal) across molecular/cell biology, biochemistry, microbiology, genetics, anatomy and physiology, neuroscience, and biostatistics. ●Designed multimodal tasks using phylogenetic trees, microscopy images, gel electro
  • Outlier
    Oracle Member – Elite Contributor
    Outlier
    Aug 2025 - Current (1 year 2 months)
    I was selected as an Oracle Member at Outlier for my exceptional performance and quality. This role gives me priority project access, opportunities to work on high-impact initiatives, and the chance to collaborate with other top contributors to deliver outstanding results.
  • Outlier
    Generative AI Tasker/ Reviewer
    Outlier
    May 2024 - Current (2 years 5 months)
    I serve as a Language Model Trainer at Scale AI Outlier, responsible for creating prompts, evaluating AI responses, and reviewing tasks for precision and consistency to maintain high-quality data for language model training.
  • B
    Research Assistant
    Biological Research on the Brain (BRB)
    May 2024 - Current (2 years 5 months)
  • Mercor
    Multilingual Call Evaluator & Transcription Contractor
    Mercor
    Apr 2026 - Jun 2026 (3 months)
    Quality assurance for Decagon's multilingual AI voice agent deployed in a financial services use case (SimplePay). ●●Key Responsibilities:●● - ●●Scenario-based call testing:●● Placed roleplay calls as a customer across predefined scenarios (transaction disputes, delayed/missing payments), simulating realistic audio environments to stress-test agent performance. - ●●Scorecard evaluation:●● Assessed each conversation against a standardized rubric covering language consistency, false language switching, turn-taking, fluency, formality, terminology localization, pronunciation, and prosody. - ●●Transcription QA:●● Performed turn-by-turn comparison of platform STT output against ground-truth speech, validating intent preservation and language-cod
  • CNTXT AI
    Data Annotator
    CNTXT AI
    Apr 2026 - May 2026 (2 months)
    ● Annotated image datasets using bounding boxes, polygons, and keypoints for computer vision models. ● Supported robotics automation and AI training workflows. ● Assisted in object detection, segmentation, and pose estimation projects. ● Maintained high annotation accuracy and quality standards. ● Worked with large-scale image data for machine learning applications.
  • Mercor
    Tool Use Expert
    Mercor
    Jan 2026 - Apr 2026 (4 months)
    Worked on AI evaluation workflows within Meta's SRT platform, focusing on improving large language model tool-calling capabilities, multi-turn reasoning, and structured response reliability. ●●Key Responsibilities:●● ●Designed multi-turn prompt scenarios to evaluate and improve tool-calling behavior in LLM systems ●Created and refined evaluation rubrics for structured reasoning, tool selection accuracy, and output correctness ●Authored golden truth (gold standard) responses for benchmarking model outputs ●Built complex prompts that trigger single and multi-tool execution workflows ●Identified failure modes in tool-use reasoning, including incorrect tool selection and parameterization errors ●Improved consistency and reliability of structure
  • SME Careers
    Data Trainer
    SME Careers
    Dec 2025 - Feb 2026 (3 months)
    - Develop and Improve AI Training Content: Create or review detailed prompts in various topics and responses to guide AI learning, ensuring the models reflect a comprehensive understanding of various topics. - Optimize AI Performance: Evaluate and rank AI responses to enhance the model's accuracy in various contexts. - Ensure Model Integrity: Test AI models for potential inaccuracies or biases, validating their applicability in the respective domains.
  • Mercor
    Biology Expert
    Mercor
    Oct 2025 - Jan 2026 (4 months)
    Worked on advanced biology-focused prompt engineering and evaluation tasks aimed at testing and improving the reasoning robustness of AI models in biomedical and life sciences domains. ●●Key Responsibilities:●● ●Designed and refined biology “stumping” prompts to test model reasoning limits in life sciences ●Evaluated AI-generated responses for scientific correctness, depth of reasoning, and consistency ●Identified failure patterns in biological reasoning (cellular processes, genetics, molecular biology concepts) ●Improved prompt quality to enhance dataset reliability for AI benchmarking and training ●Supported development of high-difficulty evaluation sets for biomedical AI performance testing
  • Mercor
    Bilingual (Bengali + English) STEM Expert
    Mercor
    Sep 2025 - Oct 2025 (2 months)
    I was responsible for auditing and refining vendor-created datasets. ●●Key Responsibilities●● ●Audited and refined large-scale AI training datasets for accuracy, consistency, and instruction adherence ●Improved multilingual STEM prompts and responses to enhance clarity, correctness, and naturalness ●Redesigned and optimized annotation rubrics for MECE-aligned evaluation workflows ●Performed detailed error analysis across multiple languages and task formats ●Strengthened dataset reliability for downstream model training and AI alignment pipelines
  • TELUS Digital
    AI Data Annotation Specialist
    TELUS Digital
    Sep 2025 - Oct 2025 (2 months)
  • Bangladesh Council of Scientific and Industrial Research (BCSIR)
    In-Plant Trainee
    Bangladesh Council of Scientific and Industrial Research (BCSIR)
    Dec 2024
  • American Society for Microbiology Student Chapter of University of Rajshahi
    Vice President
    American Society for Microbiology Student Chapter of University of Rajshahi
    Jan 2024 - Jan 2025 (1 year 1 month)
  • American Society for Microbiology Student Chapter of University of Rajshahi
    Joint Secretary
    American Society for Microbiology Student Chapter of University of Rajshahi
    Jul 2023 - Dec 2023 (6 months)
  • ABCD Laboratory, Bangladesh
    Junior Research Collaborator
    ABCD Laboratory, Bangladesh
    Feb 2023 - Feb 2024 (1 year 1 month)
  • Campus Lite
    Human Resource Executive & Content Creator
    Campus Lite
    Jun 2020 - Apr 2021 (11 months)
  • American Society for Microbiology Student Chapter of University of Rajshahi
    General Member
    American Society for Microbiology Student Chapter of University of Rajshahi
    Mar 2020 - Jun 2023 (3 years 4 months)
Education verified_user 0% verified
  • University of Rajshahi
    master, Master of Science - M.Sc. Biochemistry and Molecular Biology
    University of Rajshahi
    Jan 2025 - Jan 2026 (1 year 1 month)
  • University of Rajshahi
    bachelor, Bachelor of Science - B.Sc. Biochemistry and Molecular Biology
    University of Rajshahi
    Jan 2020 - Jan 2025 (5 years 1 month)
  • Dhaka Imperial College
    Higher Secondary School Certificate Science
    Dhaka Imperial College
    Jan 2017 - Jan 2019 (2 years 1 month)
  • N
    Secondary School Certificate Science
    National Ideal School & College
    Jan 2015 - Jan 2017 (2 years 1 month)
This is a community-created genome.