Levi M.

Key Skills

Software

AWS SageMaker

Anno-Mage

Axiom AI

Clickworker

Top Subject Matter

Conversational AI/Sentiment Analysis

Business Process Automation/Natural Language Understanding

Top Data Types

Audio

Text

Image

Top Task Types

Emotion Recognition

Prompt + Response Writing (SFT)

Segmentation

Classification

Text Generation

Text Summarization

Transcription

Evaluation/Rating

Computer Programming/Coding

Function Calling

Freelancer Overview

AI IVR Data Annotation and Model Training. Brings 4+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Bachelor of Science, The Catholic University of Eastern Africa (2026) and Certificate, ALX Africa (2024). AI-training focus includes data types such as Audio and Text and labeling workflows including Emotion Recognition and Prompt + Response Writing (SFT).

Labeling Experience

GPT Prompt-Response Data Labeling for Dynamic Approvals

TextPrompt + Response Writing (SFT)

Integrated Azure OpenAI and Copilot into enterprise applications to enable intelligent automation and AI-assisted user experiences. I was responsible for writing and curating prompt-response data for training and fine-tuning GPT models used in workflow and approval system automation. Labeled data was used to instruct and teach AI how to interpret plain English business rules and translate them into software actions. • Authored and reviewed prompt-response training pairs for workflow automation. • Evaluated and refined AI-generated outputs for accuracy in rule mapping. • Worked closely with business analysts to ensure domain relevance. • Labeled examples for both intent understanding and stepwise automation.

2025 - 2026

AI IVR Data Annotation and Model Training

AudioEmotion Recognition

Developed and deployed an automated AI IVR system utilizing IBM Granite LLM and Azure Cognitive Services for voice and sentiment analysis. I worked on building models for handling and interpreting real customer calls, including sentiment analysis and response generation. The project involved processing large sets of audio data and annotating them to train AI for detecting customer intent and emotions. • Annotated voice call datasets for intent and sentiment. • Fine-tuned models for conversational AI using Azure OpenAI tools. • Validated AI outputs and corrected intent/emotion labeling errors. • Used IBM Granite and Azure Cognitive Services for labeling tasks.

2025 - 2026

Education

Work History