The latest articles on DEV Community by Rachit Gupta (@rachit_gupta_3188854dd0a7). https://dev.to/rachit_gupta_3188854dd0a7 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F2919769%2F849e4ca8-6321-447c-8fd5-81d84b157733.png https://dev.to/rachit_gupta_3188854dd0a7 en Rachit Gupta Fri, 07 Mar 2025 06:23:00 +0000 https://dev.to/rachit_gupta_3188854dd0a7/building-adaptive-platforms-keeping-costs-in-check-njk https://dev.to/rachit_gupta_3188854dd0a7/building-adaptive-platforms-keeping-costs-in-check-njk <p>I’ve been researching how to build an adaptive coding platform, and it got me diving into similar models in other domains—like <a href="https://www.ixl.com/" rel="noopener noreferrer">IXL</a> for Math, <a href="https://www.learntheta.com/" rel="noopener noreferrer">LearnTheta</a> for Aptitude, and <a href="https://www.duolingo.com/" rel="noopener noreferrer">Duolingo</a> for language learning.</p> <p>One thing that puzzles me: if I were to use a strong LLM for every operation, the compute cost would skyrocket, especially with heavy users. So how do these platforms keep costs in check while delivering real-time adaptation? I tried some free models, and results were not great. </p> <p>Anyone with experience in this space? Would love to hear thoughts on backend optimizations!</p> adaptivelearning llm duolingo learntheta Rachit Gupta Fri, 07 Mar 2025 06:04:14 +0000 https://dev.to/rachit_gupta_3188854dd0a7/why-does-no-one-talk-about-gemini-3a64 https://dev.to/rachit_gupta_3188854dd0a7/why-does-no-one-talk-about-gemini-3a64 <p>Everyone’s hyped about GPT, Claude, and DeepSeek, but Gemini hardly gets a mention. I understand it lacks the finesse of the others. But for everyday tasks, strong language skills, and a solid free tier, it’s actually great. Google just isn’t pushing it as hard. Less hype, still useful.</p> <p>Have you tried it? What are your thoughts?</p> ai gemini llm google