Nathan Kasanda

Key Skills

Software

AWS SageMaker

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Top Label Types

Fine Tuning

Evaluation Rating

Computer Programming Coding

Function Calling

Prompt Response Writing SFT

Freelancer Overview

Drawing from my Master of Science in Computer Science from Georgia Tech and over five years of experience in AI-augmented operations, I specialize in the high-level supervision and fine-tuning of Large Language Models (LLMs). My expertise is centered on RLHF (Reinforcement Learning from Human Feedback) and Supervised Fine-Tuning (SFT) , particularly for technical domains such as Python-based autonomous agents and complex API integrations. I have a proven ability to evaluate model-generated code for algorithmic efficiency and security, ensuring that automated systems adhere to strict operational standards while maintaining high-quality outputs. What sets me apart is the unique intersection of my technical coding proficiency and my deep background in Customer Experience Management. I don't just label data; I architect the logic that powers it, as demonstrated by my project developing an autonomous routing agent using LangChain and the OpenAI API. With an AWS Certified Machine Learning – Specialty certification and a track record of reducing manual inquiries by 63% through automated knowledge base architecture , I provide the domain-specific expertise required to train AI models that are both technically robust and empathetically aligned with user needs.

Labeling Experience

RLHF & Supervised Fine-Tuning (SFT) for Python-Based Autonomous Agents

Aws SagemakerComputer Code ProgrammingFine TuningEvaluation Rating

Leveraging a Master of Science in Computer Science from Georgia Tech , I specialize in the fine-tuning and evaluation of Large Language Models (LLMs) focused on system-level programming and autonomous workflows. My expertise includes: * Code Quality Evaluation: Providing high-fidelity ratings for model-generated Python code, ensuring adherence to PEP 8 standards and optimal algorithmic complexity. * Agentic Logic Training: Developing and annotating datasets for LangChain and OpenAI API integrations to improve model performance in multi-step reasoning and autonomous ticket routing. * API & Microservices: Writing 'ground truth' SFT pairs for RESTful API integrations and microservice architectures to reduce hallucination in technical documentation tasks. * Security & Red Teaming: Evaluating model outputs for security vulnerabilities in cloud environments (AWS/Azure) , specifically focusing on preventing injection attacks in automated support systems.

Leveraging a Master of Science in Computer Science from Georgia Tech , I specialize in the fine-tuning and evaluation of Large Language Models (LLMs) focused on system-level programming and autonomous workflows. My expertise includes: * Code Quality Evaluation: Providing high-fidelity ratings for model-generated Python code, ensuring adherence to PEP 8 standards and optimal algorithmic complexity. * Agentic Logic Training: Developing and annotating datasets for LangChain and OpenAI API integrations to improve model performance in multi-step reasoning and autonomous ticket routing. * API & Microservices: Writing 'ground truth' SFT pairs for RESTful API integrations and microservice architectures to reduce hallucination in technical documentation tasks. * Security & Red Teaming: Evaluating model outputs for security vulnerabilities in cloud environments (AWS/Azure) , specifically focusing on preventing injection attacks in automated support systems.

2017 - 2021

Education

Work History