The overall goal of the assignment is to develop the A-Search platform, an AI-powered system designed to optimize research processes for the African Population and Health Research Center (APHRC). This platform will leverage existing datasets from INSPIRE and HDSS to achieve the following objectives:
Enhance research efficiency by using AI to analyze existing datasets and identify research questions that can be answered without additional data collection, saving time and resources.
Automate routine tasks by generating follow-up data collection notifications for HDSS sites, ensuring timely and consistent data collection based on pre-defined protocols and AI-driven analytics.
Improve data integration by linking records from HDSS, health facilities, and households to create a unified and comprehensive dataset, enabling more robust and holistic analysis.
Streamline data entry and transcription by integrating voice-to-text functionality, particularly in field research settings, reducing manual effort and improving accuracy.
Ensure data quality by using AI to detect anomalies in datasets, such as outliers, missing values, or inconsistencies, ensuring high data reliability for research purposes.
Empower researchers by automating repetitive tasks and providing advanced analytical tools, allowing them to focus on generating actionable insights and addressing critical health and development challenges in Africa.
Enable HDSS sites appreciate and use novel AI tools for health communication or other uses such as record linkage, data quality assessment, and record reconciliation in HDSS survey rounds.
Specific tasks
The consultant will collaborate with the DSP and IT teams at APHRC to achieve the following:
Task 1: Needs assessment and stakeholder engagement
Conduct a comprehensive needs assessment to understand the requirements of the INSPIRE Network and its stakeholders.
Engage with HDSS site managers, health facility teams, and data analysts to gather input on platform functionality and usability.
Develop a detailed system design and roadmap based on feedback and requirements.
Task 2: Conduct data mapping at the HDSS sites to assess the data ecosystem, data capability and data maturity
Determine the type of data available at each HDSS site.
Develop an AI algorithm to map the types of data collected at different HDSS sites over time, categorized by year or period since each site’s inception.
Special attention should be given to identifying family planning (FP) data.
This mapping should provide a clearer understanding of how data collection has evolved at each site.
Task 3: AI system development
AI for research question identification:
Design algorithms to scan and analyze datasets across the INSPIRE Network.
Use AI to identify trends, patterns, and gaps in the data that suggest research questions.
Develop a user-friendly interface to present suggested research questions.
AI record linkage
Build algorithms to link records from HDSS sites, health facilities, and households.
Implement advanced matching techniques to ensure data accuracy and avoid duplication.
Task 4: Automated notification system
Develop a notification system that uses pre-defined protocols and AI-driven analytics to alert HDSS sites of upcoming or overdue follow-up data collection activities. The system should be able to send alerts to HDSS sites via email, SMS, or platform notifications.
Ensure notifications are customizable based on site-specific protocols and timelines.
Task 5: Voice-to-Text integration
Build a system that supports voice-to-text to data collection to be tested in selected HDSS.
Incorporate a voice-to-text feature to transcribe audio data from interviews, focus group discussions, and other HDSS activities.
Task 6: AI-based anomaly detection
Design machine learning models to detect anomalies in datasets, including missing data, outliers, and inconsistencies during data collection.
Implement automated alerts to flag anomalies for investigation and resolution.
Task 7: Platform development and testing
Develop the A-Search platform, integrating all AI-powered features into a single, scalable system.
Conduct rigorous testing to ensure functionality, security, and scalability.
Address bugs and performance issues before deployment.
Task 8: Deployment and user training
Deploy the A-Search platform across the INSPIRE Network.
Develop training materials and deliver hands-on training sessions for researchers, data managers, and other stakeholders.
Deliverables
The consultant is expected to deliver the following:
Needs assessment report: Detailed documentation of stakeholder requirements and platform design.
Data mapping report indicating available data sets by year or period of data collection for each HDSS site.
AI Algorithms: Fully developed algorithms for research question identification, record linkage, and notifications for HDSS follow-up.
Voice-to-text application.
Anomaly detection tool.
Platform prototype: A functional prototype for testing and feedback.
Final Platform: Fully operational and deployed A-Search platform with all integrated features.
User training materials: Manuals, guides, and video tutorials for end-users.
Performance report: Evaluation of the platform’s functionality, accuracy, and user feedback.
Application documents should include:Application processApplications proposals should be submitted via email to consultancies@aphrc.org copying procurement@aphrc.org with the email subject “AI-Powered Research Optimization Platform.” Review of applications will start on February 07th 2025, and incoming applications will be reviewed on a rolling basis until the right firm is identified.
Apply via :
consultancies@aphrc.org
