Ario D.

Key Skills

Software

Other

Top Subject Matter

Multi-subject written content (World History, International Relations, Economics, Geography, Science, Governance) and multilingual text evaluation (Bengali/English)

Financial compliance knowledge for AML/KYC/CTF (sanctions, PEP screening, FATF risk categories)

Legal Services & Contract Review

Top Data Types

Text

Document

Image

Top Task Types

Entity (NER) Classification

RLHF

Object Detection

Evaluation/Rating

Freelancer Overview

Over the past three years, I have spent 1,000+ hours independently reading, evaluating, and classifying written content across eight subject domains as part of my UPSC Civil Services preparation — one of the world's most demanding examinations with a selection rate below 0.1%. This work required me to assess large volumes of information for factual accuracy, relevance, bias, and overall quality, and to compare competing sources and rank them by reasoning strength. These are precisely the skills used in AI response evaluation, search quality rating, text classification, and RLHF tasks. Alongside this, I have studied AML, KYC, and CTF compliance frameworks, which involved applying strict rule-based classification guidelines to identify and categorise financial entities — a process closely aligned with named entity recognition and structured data labeling work. What sets me apart is my language profile and subject breadth. I am a native Bengali speaker and fluent in Hindi — two of the most in-demand languages on AI training platforms, covering over 800 million people. This makes me well suited for multilingual content evaluation, localisation review, and language-specific annotation projects. My academic background in Mathematics, Physics, and Computer Science from the University of Calcutta gives me a logical, systematic approach to any task. I type at 50+ WPM, am familiar with annotation platforms including Labelbox, CVAT, and Appen, and have a consistent record of following detailed guidelines precisely — whether in compliance study, examination frameworks, or community project work through Rotaract.

Labeling Experience

Compliance Research & Rule-Based Classification (AML, KYC & CTF Certification Study)

OtherTextEntity (NER) Classification

Self-directed study of AML, KYC, and CTF rules focused on identifying and classifying financial entities using defined regulatory criteria. The training emphasized consistent application of sanctions screening logic and structured decision trees, aligning with labeling and categorization tasks. Understanding the downstream impact of small classification errors reinforced a detail-first quality approach. • Classified financial entities using OFAC sanctions lists, PEP screening criteria, and FATF risk categories. • Applied regulatory guidelines (EU AMLD, IFSCA) consistently across different case types. • Used structured, rule-based decision frameworks to support accurate categorisation. • Emphasized error sensitivity to protect data quality in labeling workflows.

2023 - Present

Self-Directed Research & Content Evaluation (UPSC Civil Services Preparation)

OtherText

Self-directed content evaluation for UPSC preparation built practical experience in judging quality, relevance, and factual correctness across long-form written material. The work involved applying consistent evaluation frameworks and criteria without deviation, similar to guideline-based annotation. It also included multilingual content review and speed-focused transcription-like writing under time constraints. • Read and evaluated 1,000+ hours of written content across eight subject areas. • Compared multiple sources to decide which explanations were more accurate and well-reasoned. • Identified factual errors, missing context, and one-sided framing for moderation/classification-style tasks. • Practised Bengali-English bilingual assessment while maintaining consistency and accuracy.

2020 - Present

Education

Work History