LLM Evaluation · Multimodal AI · Computer Vision · CUA · Automation
I evaluate AI systems, build high-quality training data, and develop automation tools for complex data workflows.
I work at the intersection of AI evaluation, training data, computer vision, and software engineering.
My work includes evaluating LLM and multimodal outputs, reviewing computer-use agent trajectories, comparing model responses, validating structured datasets, and performing image and video annotation for computer vision systems.
I also build software and automation around data-intensive workflows. My engineering stack includes Python, C, JavaScript, TypeScript, Node.js, SQL, MongoDB, Linux, and REST APIs, with experience developing browser automation, scrapers, bots, data-processing utilities, and validation tools.
The common thread across my work is AI data quality: identifying errors, enforcing evaluation criteria, validating outputs, and building tools that make complex workflows more reliable and repeatable.
LLM Evaluation Generative AI Multimodal Evaluation SxS Evaluation CUA Evaluation AI Training Data Data Annotation Quality Assurance
Evaluation of model outputs for factuality, instruction following, completeness, grounding, tool use, task completion, and overall response quality.
CVAT Computer Vision Annotation Video Object Tracking Image Annotation Video Annotation Bounding Boxes Segmentation Visual QA
Annotation and quality evaluation of visual datasets, including object tracking across video sequences and structured annotation workflows.
Puppeteer Web Scraping Browser Automation REST APIs Telegram API Postman ETL
Development of repeatable workflows for data collection, processing, validation, transformation, and API-driven automation.
Languages
Python C JavaScript TypeScript SQL Dart
Backend & Automation
Node.js Puppeteer REST APIs Web Scraping Browser Automation Telegram API
Data
MongoDB SQLite JSON CSV ETL
AI & Computer Vision
LLM Evaluation Multimodal AI CVAT Video Object Tracking Data Annotation Dataset QA
Systems & Development
Linux Bash Git GitHub Docker Postman
Additional Tools
Flutter Inkscape
A Python-based evaluation framework for benchmarking LLM and multimodal AI outputs across factuality, instruction following, completeness, grounding, tool use, and response quality.
The project includes structured evaluation schemas, pairwise response comparison, error classification, scoring utilities, and evaluation reporting.
Stack: Python JSON AI Evaluation Multimodal AI
Status: In development
A Python-based media downloading utility focused on practical automation and streamlined media-processing workflows.
Stack: Python Automation
A browser extension extending media workflow functionality directly into the browser.
Stack: JavaScript Browser APIs Automation
A C project exploring binary space partitioning and low-level graphical programming.
Stack: C raylib Algorithms
A reproducible CVAT-based project demonstrating object tracking, track continuity, occlusion handling, annotation QA, structured exports, and visualization of tracked objects.
CVAT Python OpenCV Computer Vision
A Python toolkit for detecting malformed annotations, invalid bounding boxes, missing labels, duplicate annotations, class inconsistencies, and dataset anomalies.
Python Computer Vision COCO YOLO Data Quality
A TypeScript and Node.js pipeline for automated browser interaction, structured data extraction, validation, normalization, storage, and reporting.
TypeScript Node.js Puppeteer MongoDB
I am currently developing open-source projects around AI evaluation infrastructure, multimodal data quality, computer vision annotation, and workflow automation.
My current work focuses on:
- Structured LLM and multimodal evaluation
- Pairwise and side-by-side model comparison
- Computer-use agent evaluation
- Evaluation error taxonomies
- Computer vision annotation quality
- Video Object Tracking
- Dataset validation and QA
- Browser and API automation
- Python-based data tooling
I am interested in engineering and data roles involving:
AI Evaluation · LLM Training · AI Data Quality · Computer Vision · Multimodal AI · CUA Evaluation · Data Annotation · Automation · Python Development
AI Evaluation · Computer Vision · Automation · Software Engineering