ayoub jridi@ayoubjridi

Proload is a modern, AI-enhanced social networking platform built specifically for athletes and fitness enthusiasts. It offers a rich set of features that go beyond typical workout trackers or generic social platforms by blending intelligent training recommendations with engaging community tools. The platform allows users to create and share custom workouts, track their fitness progress through detailed dashboards, and connect with like-minded athletes. To power its intelligent features, Proload integrates two AI engines: TinyLlama: a lightweight, locally deployed LLM that provides private, on-device workout suggestions and fitness insights based on user context, goals, and history. Groq API + LLaMA 3.1: a high-speed, cloud-based LLM integration used for generating structured, detailed workout plans from natural language prompts (e.g., "4-day hypertrophy routine for home with dumbbells"). This hybrid AI architecture ensures both performance and privacy: local models handle lightweight inference, while Groq’s LLaMA 3.1 adds depth and sophistication when needed. Built with Next.js and NeonDB, the platform uses a monolithic yet scalable architecture, combining AI-generated content with interactive features such as workout cloning, social posts, reactions, comments, and personalized recommendations. By combining the privacy of TinyLlama with the power of Groq’s LLaMA 3.1, Proload offers an adaptive, motivating, and data-driven experience that empowers users to reach their fitness goals efficiently and intelligently.