MamaMode

Inspiration

As university students, we understand what it’s like to feel stressed, overwhelmed, or isolated. Between late night study sessions, balancing assignments, and navigating personal challenges, it can become difficult to find someone to talk to. We realized that many people, specifically new mothers, could benefit from an always-available, supportive companion who listens, offers guidance, and provides comfort anytime they need it, especially when navigating a new lifestyle with a baby. That’s what inspired MamaMode, a nurturing AI that provides empathetic support, helping new mothers feel understood and provide support at any time.

What it does

MamaMode is a 24/7, phone-first support companion for new parents through postpartum. Instead of requiring an app or lengthy setup, MamaMode meets parents where they already are, through simple SMS or voice calls. When exhaustion and overwhelm hit, especially in the middle of the night, users can text or call and receive warm, human-sounding guidance with one small, doable next step. The assistant uses intelligence to determine whether a parent needs comfort, practical resources, or escalation to professional care, and she responds without judgment. She remembers key context, such as the baby’s age, feeding method, recovery status, and prior conversations, allowing her to tailor advice over time and create a sense of continuity. Designed specifically for low-bandwidth moments, MamaMode focuses on tiny, actionable guidance rather than overwhelming information, making support feel accessible and realistic.

How we built it

MamaMode was built by using various technologies. Google Gemini created the thoughtful, empathetic responses and Eleven Labs brought the experience to life with realistic voices. Twilio API enabled the SMS and voice communication, allowing the new mothers to interact with the app naturally. All conversations and user data were securely stored and personalized using the Snowflake API, so the AI could remember the context and provide more tailored support over time.

Challenges we ran into

While integrating Snowflake as our data storage component, we faced a security related obstacle. The default configuration only allowed connections from a specific IP address, otherwise returning a 401 Unauthorized error. This resulted in us creating a custom privacy policy and configuring access permissions to ensure our application could securely communicate with Snowflake without running into issues with sensitive user data.

Accomplishments that we're proud of

We’re most proud that we built something emotionally meaningful under intense time pressure. Instead of just connecting APIs, we were able to create a system that reacts with warmth, safety and intention in moments where parents feel most alone. We’re also proud that we expanded beyond our original scope and successfully implemented both SMS and voice calls, making Auntie more accessible to exhausted parents who may not have the energy to type. Most importantly, we’re proud that this project feels genuinely human and not just technically impressive, and has real potential to lower the barrier to postpartum support in vulnerable moments.

What we learned

Through building MamaMode, we learned that creating responsible AI for sensitive use cases is very different from building a typical chatbot. It required us to think carefully about tone, safety, escalation, and the real emotional impact of every response. We learned how important it is not to just generate text, but categorize intent and risk in a way that allows the system to respond appropriately. Technically, we gained hands-on experience integrating multiple APIs under tight time constraints, troubleshooting web hook flows for SMS and voice, and designing a system where memory, real-time processing, and AI reasoning all work together seamlessly.

What's next for MamaMode

We hope to make MamaMode more accessible by integrating it with WhatsApp, allowing users to interact with the AI through a platform they may already use daily. Additionally, we hope to incorporate multiple languages, allowing users from different backgrounds to communicate naturally and comfortably with the app, making the experience more personalized.

Built With

• elevenlabs
• gemini
• mongodb
• ngrok
• snowflake
• twilio
• typescript