I am looking to hire a skilled Machine Learning + Computer Vision developer to build a cricket pitch analytics system similar to FullTrack.ai. The system must detect pitch regions, classify ball lengths, track ball trajectory, and generate next-ball predictions. Project Requirements (Software Part) Pitch Training & Automatic Length Zoning Train a model to identify a cricket pitch from video frames. Automatically divide the pitch into four fixed length zones based on distance from the striker stumps: Yorker Full Good Short Zones must be dynamically detected using the ground markings / pitch reference points. Ball Pitch Detection & Length Classification From every ball video: Track the ball until it bounces. Detect the exact pitch location. Classify it as yorker/full/good/short. Store ball-by-ball pitch length results in a database / CSV. Next-Ball Intelligent Suggestion (Basic ML Training) System should learn from the previous ball length. Next ball should be suggested such that: If the previous ball was Good, next must be (Yorker / Full / Short) — Good excluded. Random selection among remaining 3 zones. Simple reinforcement-like logic is enough (no heavy RL needed). Dashboard / Visualization The final dashboard must display: A pitchmap like the reference (dots on zones). Ball tracking path (frame-by-frame or simplified visualization). Summary of all bowled balls and their classified lengths. Next-ball predicted length suggestion. Reference System https://www.fulltrack.ai/ (Use only for understanding the expected pitchmap output style.)