Independent Research Partner – Comprehensive Evaluation of the WAZZII Digital Health Chatbot in Kenya – Drive Impact with brighthush

About brighthush and the Digital Health Revolution

brighthush is a global pioneer in harnessing technology to empower young people—especially girls and young women—to make informed, confident health decisions. Rooted in youth‑centred design and evidence‑based behaviour change, brighthush has built a portfolio of digital platforms that deliver trusted sexual and reproductive health (SRH) information, maternal and newborn health (MNH) support, and seamless pathways to real‑world services. Our work is guided by an unwavering commitment to equity, safety, and measurable impact, ensuring that every user—regardless of geography, language, or digital literacy—receives the support they need.

In Kenya, brighthush is rolling out three transformative programmes that integrate the WAZZII chatbot, a cutting‑edge AI‑driven conversational agent, with multimedia content, community mobilisation, and direct service linkages. These programmes aim to close critical gaps in SRH and MNH knowledge, improve health outcomes, and create scalable models for digital health delivery across sub‑Saharan Africa.

Programmes in Focus

• MSD 4.0 – Nairobi & Nakuru: Enhancing maternal and newborn health through targeted messaging and real‑time support.
• Tukisonga 2.0 – Nairobi & Nakuru: Delivering comprehensive SRH education to adolescent girls and young women (AGYW) aged 15–24, covering contraception, DMPA‑SC/self‑injection, heavy menstrual bleeding, and STI prevention.
• AI Mobiliser – Nairobi: Leveraging AI to provide personalised SRH information, facilitate informed decision‑making, and streamline progression to contraceptive services.

Why This Evaluation Matters

As digital health solutions scale, rigorous evaluation is essential to understand their real‑world impact, identify areas for improvement, and ensure safety and equity. The WAZZII evaluation will provide brighthush with actionable insights into:

• How users engage with the chatbot across different contexts.
• AI performance, including accuracy, reliability, and safety in handling sensitive topics.
• Behavioural outcomes such as knowledge gains, confidence, and service uptake.
• Equity of access and outcomes across diverse user groups.
• Cost‑efficiency and potential for scaling the model.

Key Responsibilities

As the independent research partner, you will lead a comprehensive, mixed‑methods evaluation that spans data analytics, AI testing, user research, and service pathway analysis. Your work will directly influence the next phase of brighthush’s digital health strategy and inform global best practices.

Evaluation Design & Analytics

• Develop a robust evaluation protocol, indicator framework, sampling strategy, and analysis plan that aligns with brighthush’s strategic objectives.
• Conduct a deep dive into platform analytics to profile user characteristics, engagement patterns, repeat usage, and drop‑off points.
• Map the journey from programme exposure to chatbot interaction, human support, referral, and final service uptake.
• Integrate platform data with user surveys, qualitative research, and existing referral/service datasets to create a holistic evidence base.
• Identify data gaps, selection bias, and limitations in attributing outcomes to chatbot use, and propose mitigation strategies.

AI Performance, Safety & User Experience

• Assess WAZZII’s accuracy, reliability, relevance, comprehension, accessibility, privacy, and trustworthiness across all supported languages.
• For AI Mobiliser, conduct structured testing of representative SRH queries, including contraception, DMPA‑SC/self‑injection, sensitive disclosures, misinformation risks, and referral pathways.
• Review AI performance monitoring, safeguarding alerts, failure logs, and escalation procedures to ensure continuous improvement.
• Apply the 22‑metric Product‑Level Scorecard to benchmark AI safety and performance against industry standards.
• Use user surveys, interviews, and usability sessions to uncover barriers and opportunities for improvement.

Equity & Digital Access

• Examine how users with varying levels of digital access—shared devices, connectivity, digital literacy, privacy concerns, language, disability, and geography—engage with WAZZII.
• Analyse disparities in engagement, experience, referral, and outcomes by user characteristics.
• Explore interactions between chatbot use, multimedia content, community mobilisation, and human support, especially within AI Mobiliser.

Behavioural Outcomes & Service Pathways

• Measure reported knowledge, confidence, decision‑making, intentions, and service‑seeking behaviours using programme‑relevant metrics.
• For AI Mobiliser, track progression along the contraceptive‑care pathway, including referrals offered, completed visits, and service uptake.
• Utilise referral codes, provider records, voucher redemption, or other verified methods to confirm service uptake.
• Differentiate between verified and self‑reported uptake, ensuring rigorous attribution.
• Where feasible, assess outcomes by engagement intensity and combinations of programme exposure, while maintaining methodological rigor to avoid over‑interpreting associations as causal effects.

Analysis & Recommendations

• Triangulate findings across analytics, AI testing, user research, and service data to build a comprehensive evidence narrative.
• Prioritise actionable recommendations for AI safety, product design, accessibility, referral systems, and programme implementation.
• Prepare clear, evidence‑based insights that can be communicated to stakeholders, funders, and the broader digital health community.

Deliverables

• Inception report, evaluation protocol, and indicator framework.
• Analytics review and user‑journey map, highlighting data gaps.
• User research tools and AI performance & safety assessment framework.
• Ethics approvals and research authorisations.
• Clean, de‑identified datasets, codebooks, and analysis syntax (subject to data governance).
• Draft and final evaluation reports covering user experience, AI performance, safety, outcomes, equity, and service uptake.
• Prioritised recommendations and a presentation of findings to brighthush.

Timeline & Management

The evaluation will commence immediately upon contract award and must deliver all final deliverables by 31 December 2026. The agency will report to brighthush’s designated Evidence and Insights focal point and collaborate closely with product, programme, safeguarding, operations, and country teams. Any risks to scope, quality, budget, or deadline must be reported promptly with a mitigation plan.

Who You Are

We are looking for a research partner with deep expertise in digital health or AI evaluation, mixed‑methods research, usability testing, digital analytics, AI safety assessment, SRH/MNH, safeguarding, and research ethics. Experience conducting research in Kenya, measuring referral and service uptake, and a proven track record of delivering high‑quality evaluation reports are essential.

Essential Qualifications

• Demonstrated experience in digital health or AI evaluation.
• Proficiency in mixed‑methods research, including quantitative analytics and qualitative inquiry.
• Expertise in AI safety assessment and safeguarding protocols.
• Strong background in SRH/MNH research and measuring service uptake.
• Experience conducting research in Kenya or similar contexts.

Preferred Qualifications

• Previous work with chatbot or AI‑driven health platforms.
• Knowledge of Kenyan health system and referral pathways.
• Experience with ethics approvals and data governance in low‑resource settings.
• Capacity to manage a multi‑disciplinary research team.

Compensation, Perks & Benefits

While specific compensation details will be negotiated, brighthush offers a competitive fixed‑price contract aligned with the scope of work. Additional benefits include:

• Opportunity to contribute to high‑impact digital health solutions that improve lives.
• Collaborative partnership with a global leader in youth‑centred health innovation.
• Access to a network of experts in AI, digital health, and public health research.
• Potential for future engagements and long‑term collaboration.

How to Apply

Interested agencies should submit separate technical and financial proposals. The technical proposal must not exceed 10 pages and include CVs, references, company documents, and supporting evidence. The financial proposal should be in Kenyan shillings and detail a fixed‑price budget by phase and deliverable, personnel roles, professional rates, taxes, and reimbursable expenses.

Both proposals should be emailed as separate attachments to: [email protected] with the subject line: WAZZII Chatbot Evaluation – Kenya. Clarification questions must be submitted by the clarification deadline (19 October 2026). All proposals must be received by 26 October 2026. Late or non‑compliant submissions will not be evaluated.

Evaluation Criteria

Proposals will be assessed in the following order:

• Pre‑screening
• Technical evaluation (70 %)
• Financial evaluation (30 %)

Technical evaluation will consider:

• Understanding of WAZZII and programme requirements (10 %)
• Evaluation design, sampling, outcomes, and service pathway methodology (15 %)
• Product analytics, AI performance, and safety‑testing approach (15 %)
• User experience, equity, ethics, safeguarding, and data‑protection approach (10 %)
• Relevant organisational experience and referees (10 %)
• Team capability and work plan feasibility (10 %)

Financial proposals will be judged on value for money and overall cost. The lowest‑priced proposal will not automatically win; the best combination of technical excellence and financial viability will be selected.

General and Commercial Terms

• Contracting and Payment: The successful agency will sign brighthush’s applicable agreement. Payments will be linked to satisfactory delivery and written acceptance of agreed milestones.
• Pricing, Taxes and Expenses: Prices must be quoted in Kenyan shillings and remain firm for the proposal‑validity period. Taxes, licences, participant costs, third‑party costs, and approved reimbursable expenses must be itemised separately.
• Research Data and Intellectual Property: Subject to the final agreement, brighthush will own all commissioned tools, test sets, scoring frameworks, datasets, codebooks, analysis syntax, reports, and other outputs. The agency may not publish, reuse, retain, or disclose them without prior written approval.
• Confidentiality, Data Protection and Security: The agency must protect chatbot conversations, personal data, system information, and confidential materials. Approved systems, encryption, and access controls must be used, incidents reported promptly, and data securely returned or deleted as instructed.
• External AI Tools: brighthush data, chatbot conversations, personal information, and confidential materials must not be uploaded to public or third‑party AI systems without prior written approval and documented security controls.
• Safeguarding, AI Safety and Research Ethics: The agency must comply with approved research and safeguarding protocols, promptly escalating high‑risk findings. Testing must not expose real users to unsafe content or interfere with live operations.

We look forward to partnering with a research team that shares our passion for evidence‑driven innovation and the empowerment of young people across Kenya and beyond.