AI Fitness Trainer Capstone Build
Budget: ₹1,500 – ₹12,500 INR
I’m wrapping up my final-year project and need an AI module that can generate truly personalized workout plans. The focus is narrow and clear: every plan must be driven by the user’s stated fitness goal of “improving overall fitness.” I’m not looking for real-time form feedback or progress analytics at this stage—just rock-solid plan generation that feels tailored, practical, and evidence-based.
Here’s the flow I have in mind. A user enters basic profile data and selects “Improve overall fitness” as a goal. Your algorithm interprets that goal, selects appropriate training modalities, adjusts intensity and volume to suit beginner through intermediate levels, and outputs a structured weekly routine (exercises, sets, reps, rest, and optional equipment notes). The system should justify its choices in plain language so the end user understands the reasoning behind each exercise block.
I’d like the core to be written in Python, preferably leveraging familiar ML libraries such as TensorFlow, PyTorch, or scikit-learn—happy to hear your recommendation. Data sources for exercise selection can come from public, reputable fitness datasets or a curated ruleset you supply. The end result should be easy for me to integrate into a simple web or mobile front end.
Deliverables
• Clean, well-commented source code for the workout-plan generation engine
• A sample dataset (or documentation of public datasets used)
• A brief technical report explaining model logic, data preprocessing, and how the plan adapts to different user profiles while staying aligned with the “overall fitness” goal
• README with setup instructions so I can demo it during my submission
Acceptance criteria
• Given at least three distinct user profiles, the system produces unique, sensible plans within seconds
• All generated plans reflect balanced cardio, strength, and mobility work suitable for general fitness improvement
• Code runs locally on a standard laptop without paid APIs
If you’ve built something similar or have a strong background in fitness-oriented recommender systems, I’d love to see your approach. Let’s get this across the finish line!
Here’s the flow I have in mind. A user enters basic profile data and selects “Improve overall fitness” as a goal. Your algorithm interprets that goal, selects appropriate training modalities, adjusts intensity and volume to suit beginner through intermediate levels, and outputs a structured weekly routine (exercises, sets, reps, rest, and optional equipment notes). The system should justify its choices in plain language so the end user understands the reasoning behind each exercise block.
I’d like the core to be written in Python, preferably leveraging familiar ML libraries such as TensorFlow, PyTorch, or scikit-learn—happy to hear your recommendation. Data sources for exercise selection can come from public, reputable fitness datasets or a curated ruleset you supply. The end result should be easy for me to integrate into a simple web or mobile front end.
Deliverables
• Clean, well-commented source code for the workout-plan generation engine
• A sample dataset (or documentation of public datasets used)
• A brief technical report explaining model logic, data preprocessing, and how the plan adapts to different user profiles while staying aligned with the “overall fitness” goal
• README with setup instructions so I can demo it during my submission
Acceptance criteria
• Given at least three distinct user profiles, the system produces unique, sensible plans within seconds
• All generated plans reflect balanced cardio, strength, and mobility work suitable for general fitness improvement
• Code runs locally on a standard laptop without paid APIs
If you’ve built something similar or have a strong background in fitness-oriented recommender systems, I’d love to see your approach. Let’s get this across the finish line!