Saraphin K.

Key Skills

Software

Other

Top Subject Matter

LLM Output Evaluation

Technical Documentation

Medical Records Management

Top Data Types

Text

Document

Top Task Types

Data Collection

Classification

Prompt + Response Writing (SFT)

Freelancer Overview

LLM Output Evaluator / AI Trainer (Javan Informatics). Brings 2+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Internal, Proprietary Tooling, and Google Workspace. Education includes Bachelor of Business Information Technology, Mount Kenya University (2022) and Professional Certificate in AI for Software Engineering, Power Learn Project (2025). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Evaluation, Rating, and Data Collection.

Labeling Experience

LLM Output Evaluator / AI Trainer (Javan Informatics)

Text

Reviewed and rated AI-generated outputs from LLMs for accuracy, relevance, and coherence using Python-based tools and LLM APIs. Evaluated model outputs against prompt design requirements, enhancing AI documentation and deployment. Automated LLM output evaluations to ensure structured, high-quality technical documentation before deployment. • Used prompt engineering techniques for quality assessment. • Feedback directly influenced LLM performance in production use cases. • Collaborated with backend and documentation teams. • Demonstrated iterative improvement in evaluation protocols.

2025 - Present

AI-Augmented Technical Documentation Evaluator (Task Automator Project)

Other

Developed and tested Python-based LLM API tools to automate technical documentation and error-log generation, with iterative evaluation of model outputs for format, accuracy, and compliance. Reviewed outputs to ensure alignment with technical requirements and labeling guidelines. Implemented automated AI evaluation systems to streamline error vetting workflows in software engineering contexts. • Applied LLM output review for technical writing automation. • Used coding expertise for annotation and post-processing QA. • Collaborated with engineering teams to improve API output quality. • Integrated AI tools in backend programming environments.

2025 - 2025

Prompt Designer & SFT Evaluator (Legal Tech AI Project)

OtherTextPrompt + Response Writing (SFT)

Designed NLP prompts and evaluated AI-generated legal assistant responses using Kenyan constitutional law as the knowledge base. Integrated NLP modules and assessed text outputs for context relevance, grammar, and legal accuracy. Demonstrated data labeling, annotation, and SFT via iterative evaluation rounds to improve response accuracy. • Crafted, tested, and refined input prompts and expected outputs. • Performed manual evaluation of model-generated responses. • Utilized Python and LLM APIs (OpenAI-compatible) for deployment. • Delivered comprehensive feedback to guide model fine-tuning.

2025 - 2025

Medical Data Accuracy/Structured Labeling Specialist (Equity Afia)

TextData Collection

Managed and maintained digital patient records with utmost accuracy and adhered to confidentiality in a high-volume medical setting. Ensured structured documentation and integrity of all patient data for consistent and reliable access. Demonstrated the importance of structured workflows and high-standard data integrity in sensitive annotation environments. • Established quality assurance and verification standards. • Processed large data volumes under strict deadlines. • Used Google Workspace tools for systematic data entry. • Translated technical processes for non-technical staff.

2025 - 2025

Incident Logging Data Annotator (KUTRRH)

TextClassification

Maintained structured incident and technical logs, pre-vetted for accuracy and categorized for escalation, directly analogous to data annotation and quality assurance workflows. Provided detailed and accurate records for complex problem-solving requirements. Demonstrated classification and structured documentation skills suitable for AI data labeling contexts. • Utilized ticketing systems for log management and categorization. • Ensured escalation judgment met senior engineering standards. • Documented technical diagnostics and incidents systematically. • Facilitated communication between technical and non-technical staff.

2024 - 2025

Education

Work History