AI Model Redevelopment for Case Identification
Budget: $750 – $1,500 USD
I'm seeking a skilled AI developer to improve and redevelop our current AI model for case identification. The current model, based on AWS Comprehend, is underperforming with an accuracy rate of less than 0.90. The new model should be able to effectively differentiate between a case our company can work on and those it can't, based purely on the text content.
Key requirements include:
- Reaching a minimum 0.90 accuracy rate.
- Considering the context of the case in the text, not just the keywords.
- Training the model with annotated text examples that clearly distinguish between relevant and irrelevant cases.
In addition, the new model should be able to make intelligent predictions based on similarities with past cases in situations where the text is ambiguous or unclear.
Ideal skills and experience for this role include:
- Proficiency in AI and machine learning, particularly in text processing and classification.
- Experience in redeveloping and improving existing ML models.
- Strong background in training models with annotated text data.
- Ability to handle ambiguous or unclear data gracefully in an ML model.
If you're confident in your ability to deliver a high-performing AI model that meets these criteria, I'd love to hear from you.
Key requirements include:
- Reaching a minimum 0.90 accuracy rate.
- Considering the context of the case in the text, not just the keywords.
- Training the model with annotated text examples that clearly distinguish between relevant and irrelevant cases.
In addition, the new model should be able to make intelligent predictions based on similarities with past cases in situations where the text is ambiguous or unclear.
Ideal skills and experience for this role include:
- Proficiency in AI and machine learning, particularly in text processing and classification.
- Experience in redeveloping and improving existing ML models.
- Strong background in training models with annotated text data.
- Ability to handle ambiguous or unclear data gracefully in an ML model.
If you're confident in your ability to deliver a high-performing AI model that meets these criteria, I'd love to hear from you.