1

CONSULTANCY TO FACILITATE THE DEVELOPMENT OF AN INTELLIGENT SEARCH:AN AI-POWERED RESEARCH OPTIMIZATION PLATFORM FOR APHRC (A-SEARCH)

Deadline:  February 7, 2025

The African Population and Health Research Center (APHRC) is a leading Africa-based, international research and policy institution headquartered in Nairobi, Kenya, with a regional office in Dakar, Senegal.

Our work contributes to the body of evidence about the critical issues in population health and wellbeing affecting Africa’s development to provide strong direction and recommendations to policy and decision-makers at different levels. Our program priorities are organized under four thematic areas: Human Development, Health and Well-being, Data Science and Evaluation and Population Dynamics and Urbanization in Africa.

Our research is implemented within these thematic areas. In addition, the Center has other programmatic areas in Research and Related Capacity Strengthening, and Policy Engagement and Communications, also aligned to the priority program areas.

In 2019, APHRC launched the Implementation Network for Sharing Population Information from Research Entities (INSPIRE) as a partnership to improve health data collection and sharing across Africa. Starting in 2025, INSPIRE will partner with 10 health and demographic surveillance system (HDSS) sites across Africa to tackle big health problems through collaboration, innovation, data standardization, harmonization, and AI integration. This two-year project aims to increase data use in HDSS sites in Africa and allow responsible use of AI for operation and analysis of data generated at the sites.

Purpose of the Assignment

To further strengthen its research capabilities and improve the utilization of its vast data assets, APHRC plans to develop an artificial intelligence (AI) powered intelligent platform. This platform will optimize research processes, enhance data integration, and improve data-driven decision-making at HDSS sites. The platform will be designed specifically for data collected from partnering HDSS sites under the INSPIRE project.

The proposed search platform will utilize datasets generated by the INSPIRE project, which integrates data from HDSS, health facilities, and household surveys. Despite the richness of these datasets, inefficiencies in data integration, analysis, and anomaly detection hinder their optimal use. To address these challenges, APHRC’s Data Science Program will collaborate with the selected consultant to develop and implement the platform while deploying other AI-supported tools.

Assignment Description

Overall

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:

1. 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.
2. 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.
3. 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.
4. Streamline data entry and transcription by integrating voice-to-text functionality, particularly in field research settings, reducing manual effort and improving accuracy.
5. 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.
6. 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.
7. 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:

1. Needs assessment report: Detailed documentation of stakeholder requirements and platform design.
2. Data mapping report indicating available data sets by year or period of data collection for each HDSS site.
3. AI Algorithms: Fully developed algorithms for research question identification, record linkage, and notifications for HDSS follow-up.
4. Voice-to-text application.
5. Anomaly detection tool.
6. Platform prototype: A functional prototype for testing and feedback.
7. Final Platform: Fully operational and deployed A-Search platform with all integrated features.
8. User training materials: Manuals, guides, and video tutorials for end-users.
9. Performance report: Evaluation of the platform’s functionality, accuracy, and user feedback.

Timelines

The consultancy is expected to be completed within 180 working days from the date of contract signing.

Budget

The budget shared by the consultant will be discussed and aligned with the available funds.

Application documents should include:

1. The consultant firm should provide a technical proposal detailing the proposed methods, tools, and a systematic approach to developing the deliverables. The proposal should include information on the incorporation of open science principles.
2. The firm should share a financial proposal indicating the costs required to develop the A-SEARCH platform according to the stipulated requirements.
3. A detailed workplan to guide project implementation
4. Cover letter describing their eligibility for the assignment and understanding of the TOR.
5. Detailed CV of the key personnel.
6. Work profile including reverse chronological list of similar projects/assignments and contact details (name, email/phone) of supervisors/clients; and
7. An all-inclusive financial proposition with the expected remuneration. It includes fees for the deliverables, travel costs and any other applicable costs.

Application process

Applications proposals should be submitted via email to [email protected] copying [email protected] with the email subject “AI-Powered Research Optimization Platform.” Review of applications will start on February 07^th 2025, and incoming applications will be reviewed on a rolling basis until the right firm is identified.

Special Notice

APHRC is an equal opportunity employer that is committed to creating a diverse and inclusive workplace. All employment decisions are made on the basis of qualifications and organizational needs. Reasonable accommodation may be provided to applicants with disabilities upon request, to support their participation in the recruitment process.

Applicants are not required to make any payments to anyone during any stage of the recruitment process.