Tsehayneh Sinishaw

Key Skills

Software

Other

Mercor

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Top Label Types

RLHF

Fine Tuning

Computer Programming Coding

Freelancer Overview

I am a software engineer with hands-on experience in AI training data, specializing in Large Language Model (LLM) training and fine-tuning for enterprise clients such as Microsoft and ServiceNow. My work has involved supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to enhance model performance, conversational accuracy, and reliability across various cognitive domains including agentic tasks, complex instruction following, and complex reasoning in coding. I have created and validated algorithmic problems, developed agent-based coding tasks, and performed rigorous code quality assurance to ensure high-quality training data for AI systems. My technical skills include Python, Django, FastAPI, and experience with collaborative tools like Git, Jira, and Slack. I am detail-oriented, analytical, and passionate about building robust datasets that drive AI innovation.

Labeling Experience

Software Engineer at Mercor

MercorComputer Code ProgrammingComputer Programming Coding

Worked as a Software Engineer: ●​ Competitive Coding Writer – Created LeetCode-style algorithmic problems (Easy to Very Hard) with optimized solutions and test cases. ●​ Agentic Code Writer – Developed multi-step, agent-based coding tasks to evaluate LLM planning and execution capabilities. ●​ Code QA Writer – Reviewed, validated, and stress-tested model-generated code for correctness, efficiency, and edge cases.

Worked as a Software Engineer: ●​ Competitive Coding Writer – Created LeetCode-style algorithmic problems (Easy to Very Hard) with optimized solutions and test cases. ●​ Agentic Code Writer – Developed multi-step, agent-based coding tasks to evaluate LLM planning and execution capabilities. ●​ Code QA Writer – Reviewed, validated, and stress-tested model-generated code for correctness, efficiency, and edge cases.

2025 - 2025

LLM Trainer

Computer Code ProgrammingFine Tuning

- SFT (ServiceNow via Turing): Performed Supervised Fine-Tuning (SFT) of Large Language Models for ServiceNow via Turing, creating high-quality training data and instruction–response pairs across Agentic Tasks (AT) and Complex Instruction Following (CIF) domains to improve instruction adherence and model accuracy.

- SFT (ServiceNow via Turing): Performed Supervised Fine-Tuning (SFT) of Large Language Models for ServiceNow via Turing, creating high-quality training data and instruction–response pairs across Agentic Tasks (AT) and Complex Instruction Following (CIF) domains to improve instruction adherence and model accuracy.

2025 - 2026

LLM Trainer

Computer Code ProgrammingRLHF

- RLHF (Microsoft via Turing): Contributed to Reinforcement Learning with Human Feedback (RLHF) projects for enterprise AI systems used by Microsoft, working through Turing. Improved LLM reasoning, response quality, and reliability through human evaluation, ranking, and feedback on model outputs.

- RLHF (Microsoft via Turing): Contributed to Reinforcement Learning with Human Feedback (RLHF) projects for enterprise AI systems used by Microsoft, working through Turing. Improved LLM reasoning, response quality, and reliability through human evaluation, ranking, and feedback on model outputs.

2024 - 2025

Full Stack Developer, LLM Trainer(RLHF, SFT) at Turing

OtherComputer Code ProgrammingRLHFFine Tuning

• Trained and fine-tuned Large Language Models (LLMs) for enterprise clients including Microsoft and ServiceNow. Worked on Supervised Fine-Tuning (SFT) and Reinforcement Learning with Human Feedback (RLHF) to enhance model performance, conversational accuracy, and overall reliability. • Worked in 3 Cognitive domains. 1. Agentic Tasks(AT) 2. Complex Instruction Following(CIF) 3. Complex Reasoning(CR) - Coding

• Trained and fine-tuned Large Language Models (LLMs) for enterprise clients including Microsoft and ServiceNow. Worked on Supervised Fine-Tuning (SFT) and Reinforcement Learning with Human Feedback (RLHF) to enhance model performance, conversational accuracy, and overall reliability. • Worked in 3 Cognitive domains. 1. Agentic Tasks(AT) 2. Complex Instruction Following(CIF) 3. Complex Reasoning(CR) - Coding

2024 - 2025

Education

Work History