This is my algorithm for assigning particular tasks to a set of AI and human agents. ** I worked with 3 datasets: ** All 3 of these datasets are autogenerated using tasks.py
- Tasks (150) - task ID, key techs, project goal, urgency, etc.
- Profiles (20) - agent ID, AI or Human, best techs, avg completion time (minutes),
- Historical Performance Metrics (500) - task ID,
The task allocation process (executed in decision_algo.py):
- User human/AI agent preference
- Does the budget and time allow for the preference to work?
- Scoring each agents potential for a task based on shared characteristics + penalization to their score when agents are assigned to too many
- Choosing the best score out of all AI/human/hybrid agent(s) and assigning them the task!