Uchenna O.

Key Skills

Software

AWS SageMaker

Clickworker

Dataloop

Google Cloud Vertex AI

OpenCV AI Kit (OAK)

Top Subject Matter

Movie scripts and narrative document information extraction

Multimodal duplicate ad detection and scene structuring

Top Data Types

Document

Image

Text

Top Task Types

Prompt + Response Writing (SFT)

Entity (NER) Classification

Text Generation

Freelancer Overview

LLM-Powered Movie Script Parser project (entity extraction and structuring). Brings 1+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Langchain, Gemini, and qdrant. Education includes major in Electronics and nanoelectronics, MIREA Russian Technological University (2026). AI-training focus includes data types such as Document and Text and labeling workflows including Entity (NER) Classification and Text Generation.

Labeling Experience

Avito ML Cup 2025 / Wink AI Challenge (project work)

TextPrompt + Response Writing (SFT)Text Generation

Designed an automated structuring approach for digital scenes using a multimodal algorithm. Applied text, image, and metadata analysis to identify duplicate ads and produce structured outputs. Focused on reducing ambiguity by aligning extracted scene information with consistent representations. • Performed multimodal analysis combining text/image/metadata signals. • Identified duplicate advertisements and structured scene information. • Prepared outputs suitable for downstream ML training or evaluation. • Used competition-style iterative refinement to improve extraction quality.

2025 - 2025

LLM-Powered Movie Script Parser

DocumentEntity (NER) Classification

Worked on an LLM-driven pipeline to extract structured entities from unstructured movie scripts. Used regex rules and one-shot prompting to derive scene, character, and metadata fields into tabular outputs. Integrated semantic search via a vector database to support natural-language retrieval of extracted narrative elements. • Converted PDF/Word script documents into structured DataFrames. • Defined entity extraction targets such as scenes, characters, and metadata. • Built semantic indexing using a vector database for later querying. • Supported downstream question-style retrieval over the structured content.

2025 - 2025

Education

Work History