AI Response Evaluation & Ranking [Freelance]
I worked on this platform for about one year as a freelance contributor. My main task was to test and compare different AI models by giving them prompts. Sometimes I was given two responses and I had to check which one was better based on whether it followed the instructions properly or not. Other times, I had to come up with my own prompts from scratch to test how the AI would respond. A unique part of my work was that I gave prompts in Hindi language as well, and then evaluated the quality of the Hindi response. After every task, I had to write a short explanation or reason to justify my choice. This helped me develop a habit of noticing small details like missing instructions, wrong tone, or incorrect facts. I followed the project guidelines strictly to make sure the feedback I provided was accurate and helpful for improving the AI models