Telemetry-Driven Software Backlog Prioritization Using Trained Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software development prioritization methods rely heavily on subjective qualitative techniques, leading to biased and sub-optimal prioritization of product backlog items (PBIs) due to reliance on heuristics and unreliable customer feedback.
Innovation Solution
A system utilizing a trained machine learning model to predict user behavior based on telemetry data, combined with a generative AI model to generate feature summaries, determines similarities between user interactions and PBIs, and prioritizes them based on data-driven impact on user retention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If subjective qualitative techniques are used for prioritization, then the process is simple and fast, but the accuracy and objectivity of prioritization deteriorates due to bias and unreliable feedback
Solution Approach 1:
The patent replaces subjective human judgment (mechanical/qualitative assessment) with automated machine learning models that objectively analyze telemetry data. The ML models process user interaction data to generate quantitative predictions about feature importance, eliminating human bias while maintaining prioritization efficiency.
Solution Approach 2:
The patent introduces machine learning models as intermediaries between raw telemetry data and prioritization decisions. These models act as mediators that transform complex, unstructured user behavior data into actionable prioritization scores, enabling objective decision-making without requiring direct human interpretation of raw data.
2Measurement precision
If telemetry data analysis and ML models are used, then prioritization accuracy improves, but system complexity increases
Solution Approach 1:
The patent employs multi-functional machine learning models that simultaneously perform multiple tasks: analyzing telemetry data, predicting user behavior, ranking features, and generating explanations. This consolidation of multiple functions into unified models reduces overall system complexity compared to using separate specialized systems for each task.
Solution Approach 2:
The system implements self-service capabilities where the ML models automatically train on incoming telemetry data and continuously improve their predictions without requiring manual intervention. The automated model training and updating processes reduce operational complexity and maintain high accuracy over time.
3Measurement precision
If comprehensive telemetry data is collected and analyzed, then user behavior prediction accuracy improves, but data processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing and filtering telemetry data before full analysis. The system identifies and extracts only the most relevant features from raw telemetry data, reducing the volume of data requiring comprehensive analysis while maintaining prediction accuracy. This preliminary filtering occurs continuously in the background.
Solution Approach 2:
The system applies partial action by focusing computational resources on analyzing only the most impactful telemetry signals and user interactions. Rather than processing every single data point with equal depth, the ML models selectively analyze high-value signals that contribute most to prediction accuracy, reducing overall processing time while maintaining effectiveness.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Systems, methods, devices, and computer readable storage media described herein provide techniques for prioritizing software development using a trained model. In an aspect, model features are determined based on analysis of user behavior with respect to a software application. A software development prioritization (SDP) system determines data associated with the model features and utilizes a generative artificial intelligence (AI) model to summarize the model features based on the determined data. The SDP system determines, based on the summaries, a similarity between software development items and the model features and prioritizes one of the software development items over another based on the determined similarities. In a further embodiment, the SDP system causes a software development task corresponding to the prioritized software development item to be performed before another software development task corresponding to a different software development item. In an aspect, model features are determined utilizing a trained machine learning model.