Reflecting on a year with UNICEF Venture Fund

In November 2024, the UNICEF Venture Fund announced its Health Systems Strengthening Cohort, selecting four ventures out of 500 submissions from 71 countries.  By supporting these companies, the UNICEF Venture Fund will gather further evidence of promising use cases to scale solutions that can best deliver results for children. The ambition is to ensure every child, regardless of ability, socioeconomic status or circumstance, receives the healthcare they are entitled to. 

The Venture Fund is proud to see Docokids graduate. Docokids offers a pediatrician-backed chatbot offering 24/7 consultation and information for parents, providing comprehensive, affordable, and empathetic support. With the support of the Venture Fund, they developed and trained their AI algorithm based on pediatricians’ feedback, measured its impact to ensure no harm is done, and tested the chatbot with both users and doctors. This new AI-driven chatbot for pediatric fever guidance in Spanish aims to enhance support for parents in remote and rural areas and improve the efficiency of the pediatrician's work. 

In this interview with the Docokids team, they reflect on their journey over the last year developing their AI-powered chatbot for pediatric fever guidance and their future to take their AI-driven chatbot for pediatric fever guidance tool to market.

What was your biggest achievement over the last year?

First Spanish-Language AI Chatbot for Pediatric Fever guidance

Over the past 12 months, our team successfully developed the first Spanish-language AI-powered chatbot for pediatric fever guidance, designed to support parents and caregivers with trustworthy, clinically validated recommendations. This marks a major step toward democratizing access to pediatric care in Latin America. We also strengthened our partnerships with UNICEF and local healthcare networks, laying the foundation for large-scale, equitable deployment. 

Where will your solution have the greatest impact in the next phase, and why?

Empowering Caregivers and their Communities

We see our greatest impact in communities with limited access to pediatric professionals, particularly in Latin America. Socially, our solution empowers caregivers to make informed decisions and avoid unnecessary emergency visits, while promoting health education. From a business perspective, the chatbot acts as an accessible entry point to the Docokids ecosystem, connecting families, institutions, and NGOs through scalable, ethical digital health tools. 

Can you describe your prototyping process and how your solution evolved over time?

From Clinical Rules to Conversations

We began with a simple, rule-based prototype focused on step-by-step clinical logic. However, as we incorporated user feedback and medical validation, we transitioned to a dynamic graph-based structure using LangGraph, which allowed for more natural and context-aware conversations. Each iteration balanced clinical precision with empathy and cultural adaptation, resulting in a model that feels both medically reliable and human-centered. 

Can you share a memorable user or field test and the key lessons you learned?

Evolving from Speed and Logic to Empathy and Connection

We had imagined anxious parents who wanted quick answers from the chatbot in the middle of the night. However during the testing, we learnt that many users actually wanted empathy, reassurance, and thoughtful explanations, and not simple and quick instructions. A simple instruction like “go to the hospital” was not enough: they wanted to feel understood and supported, and to understand the reasoning behind each recommendation.

For this reason, we trained the chatbot to begin with more empathetic responses, acknowledging the user’s emotional state with phrases like “I know this can feel really stressful.” We also added educational explanations such as: “While most fevers are not dangerous, a temperature above 38°C in a 3-month-old baby can point to more serious underlying causes, which is why we recommend seeking medical care.”

Beyond educating, the chatbot also plays a preventive role by guiding users to emergency services when necessary. And because going to the ER can be overwhelming, it now provides practical, supportive suggestions, explaining what they can expect during the visit and reminding them to bring helpful items like a phone charger, water, snacks, or a light blanket to make the experience more manageable.

How has being Open Source benefited your solution and your company? Can you cite specific examples?

The Open Source Advantage: Transparency, Collaboration and Trust

Being open source has strengthened transparency, collaboration, and trust in our work. It opened the doors to external contributors, from medical professionals to data scientists, to review, improve, and validate our conversational logic. Sharing our methods publicly also shows Docokids commitment to open digital health innovation.

Who are the key collaborators you’re seeking, and how can they add value to your business?

Scaling Impact Through Partnerships

We are eager to collaborate with public health institutions, NGOs, and universities across Latin America that share our vision of accessible, evidence-based pediatric care. Partnerships with medical associations and AI research labs would also help strengthen our validation processes and open new opportunities for data-driven health innovation. Specifically we are making a robust plan to collaborate with Operation Smile to help reach families that need us the most. We are also working closely with Clouxter and Amazon Web Services to support our innovation.

What are you most excited about for your company next year, and what are your main goals?

The Road Ahead: Reaching Thousands of Families in 2026 and Launching New Use Case

Next year, we aim to scale our solution to thousands of families across Colombia and beyond, publish our first anonymized pediatric dataset, and strengthen our open-source community. We’re also looking forward to beginning adapting our technology for new use cases, including LactaBot AI, a version focused on breastfeeding and early childhood guidance in Latin America.

How has the UNICEF Venture Fund supported your solution beyond financing?

The UNICEF Catalyst: Aligning with Global Standards

Connect with Docokids

Interested in learning more about Docokids? Visit their website or follow them on Instagram and LinkedIn. Get in contact with the team directly via email at  [email protected].