Hire this AI Trainer
Sign in or create an account to invite AI Trainers to your job.
Clickworker
CVAT
OneForma
Text
Image
Audio
Video
Document
Classification
Bounding Box
Object Detection
Polygon
Point/Key Point
Cuboid
Transcription
Segmentation
Entity (NER) Classification
Polyline
RLHF
Evaluation/Rating
Text Summarization
Fine-tuning
Data Collection
Function Calling15+ years of hands-on experience in writing, editing, evaluation and quality assurance of various data across financial, crypto, SEO, marketing, and legal domains. These processes include classification and categorization, rule-based workflows and guideline adherence, fact-checking and research. Trilingual (English/Ukrainian/Russian) with daily communication and translation experience across technical, financial, and other materials. Daily testing of major proprietary and open-source LLMs/GenAI tools for 2 years with constant handling of hallucinations, biases, errors, and non-relevant outputs. Using ethical judgment, an analytical mindset, and domain-specific expertise gained from a diverse professional background. Engineering prompts as reasoning frameworks with in-context learning and sequential analysis for trading data, research, and complex tasks. Utilizing multiple LLMs as separate agents to cross-validate and refine responses.
ImageEvaluation/RatingWorked with visual data interpretation using AI. I routinely fed trading candlestick charts into LLMs to evaluate how well the models could read graphical patterns, interpret technical indicators, and output structured reasoning based on visual inputs.
TextEvaluation/RatingLeveraged different LLMs almost daily for past 2 years during my trading and educational workflows. I designed complex system prompts with assigned roles and strict rules for intricate tasks like searching for market inefficiencies and tabulated trading data analysis. Used AI to break down the mathematical logic of various complex trading indicators evaluating the models` ability to explain math and logic without making things up. Applied a self-consistency approach by generating several outputs for the same query across different models to find the most accurate and unbiased result.
TextData CollectionCompleted highly specific, structured tasks across 30+ crypto projects and early-stage blockchain testnets. It was a repetitive, rule-based execution process similar to crowdsourced labeling workflows. Also involved tracking and categorizing raw on-chain metrics using dedicated screeners and tools to sort data into actionable categories.
ImageEvaluation/RatingEvaluated and categorized over 1,000 ad creatives across various traffic networks. This involved constant comparative ranking (A/B testing) to see which text and image combinations hit the target metrics. I had to maintain structured Black/White lists, filter out low-quality traffic sources, and categorize campaigns based on performance data.
TextEvaluation/RatingManaged a team of writers and essentially ran a human-in-the-loop QA procedure. Processed a multitude of articles, product descriptions, tutorials, and reviews across different topics. I had to categorize human-written content, check against strict SEO guidelines, verify facts, and identify logical errors. It was a high-volume structured evaluation workflow where I acted as the final filter to ensure text outputs matched the initial brief. I also classified and categorized a large volume of texts for future publication on clients` sites.
Specialist (Master`s) degree in Business Law, Legal Support of Business
Certificate in English Language Studies, English Language
AI Prompt Engineer & LLM Tester
Trader & Market Analyst
