Create and update lists
Create Picnic Trip with sandwiches, apples, and water, then add sunscreen in a follow-up request.
Updated
Juggle Ease Voice Assistant
We added a multilingual realtime voice layer to an existing family productivity app, connecting natural-language speech to lists, todos, reminders, trackers, and events.
Voice AssistantApp Integration
View Live Website
Watch Easy organise a family picnic
Create a picnic list, add sunscreen, and save a packing task by voice. Then switch to Croatian to add juice to the same list. English subtitles are included.
1 min 58 sec
2 months
from kickoff to delivery
5
task categories controlled by voice
The Starting Point
Juggle Ease already helped families manage lists, todos, calendars, reminders, and trackers. Every action still depended on tapping, typing, and navigating the mobile interface, which made quick updates harder when a user was busy or away from the screen.
The Assistant Had to Do Work, Not Just Talk
A useful assistant had to understand a natural request, map it to the correct app action, handle mobile audio reliably, and confirm what changed. A fluent response without a dependable action would not improve the product.
Intent to action
Spoken language had to map reliably to create, edit, and delete operations across different task types.
Realtime mobile audio
Microphone access, playback, interruptions, and device behavior had to work across Android and iOS.
Multilingual interaction
The conversation needed to support major languages available through the realtime model without overclaiming perfect coverage.
Our Role
We joined the existing product team specifically for the AI voice layer. Our scope covered realtime conversation, speech handling, intent and action mapping, integration support, and the mobile edge cases around microphone and playback behavior.
Design the voice experience
We shaped the interaction so users could speak naturally and still understand what the app was about to do.
Map language to product actions
We connected user intent to the existing API rather than creating a disconnected conversational feature.
Support mobile reliability
We addressed the Android and iOS audio behavior required for the assistant to feel native inside the app.
Everyday tasks, handled through conversation
Easy connects a spoken request to the part of Juggle Ease that needs to change.
Save a task
Ask for a packing task and choose whether it needs a due date or reminder. The task appears in the existing to-do view.
Continue in Croatian
Switch languages, ask Easy to add juice, and find the new item saved as “sok” on the same list.
From Spoken Request to In-App Action
The assistant keeps the conversation and the underlying product action in the same realtime loop.
1. Listen
Speech-to-text captures the user's request through the mobile voice interface.
2. Understand the intent
The assistant identifies the task category, requested change, and information needed to complete it.
3. Execute through the existing API
The product creates, edits, or deletes the relevant list, todo, reminder, tracker, or event.
4. Confirm by voice
Text-to-speech closes the loop so the user knows what happened without returning to manual navigation.
The Product Behind the Conversation
The voice layer was built to extend an existing application rather than replace the product or the team already behind it.
Existing API action layer
The assistant uses Juggle Ease's existing API to make real changes inside the product.
Realtime speech loop
Speech-to-text, the realtime model, and text-to-speech keep the interaction responsive.
Cross-platform audio handling
The integration accounts for microphone and playback differences across Android and iOS.
Multilingual conversation
The feature supports major languages available through GPT Realtime 2.
The Outcome
The voice assistant launched after roughly two months and was ready on the promised date. It connected realtime multilingual speech to five categories of existing product actions. In the recorded demonstration, Easy creates a picnic list and a packing task in English, then responds in Croatian and adds juice to the same list.
What This Proves
Embedding AI into an existing product works best when the specialist team respects the product boundary, connects to real actions, and measures success against the feature it was responsible for delivering.
Adding AI to an existing product?
We can build the AI workflow, connect it to your existing APIs, and work alongside the product team already in place.
Product walkthrough
Inside the product
1 / 3
01
Create Picnic Trip with sandwiches, apples, and water, then add sunscreen through a follow-up voice request.
Client video testimonial
Amrita Ghosh, Founder, Juggle Ease shares their experience working with Adamant Code.
A Full AI Voice Assistant, Delivered Right on Time
Amrita Ghosh, Founder, Juggle Ease
What Amrita says
The founder on working with Adamant Code.
“You guys were ready right on time, the day you promised.”
Amrita
Founder and CEO, Juggle Ease
Have an AI product idea you can't build alone?
One strategy call. Bring the idea and we'll tell you honestly what it takes to build it.
Book a Strategy Call
Other case studies
GTS Innovative AI-Assisted Marketplace
We translated a founder's market vision into 50+ screens, 15+ workflows, and a structured invention process for inventors, service providers, investors, and admins.
Full AI product buildMarketplace / Innovation
GTS Innovative had a clear market vision and rough wireframes, but the user journeys, permissions, product logic, and AI workflow still needed definition. We turned that starting point into a complete marketplace platform.
Leelou AI Coach for TNM Coaching
95% of onboarded regular users completed at least one session.
AI coaching productCoaching / Training
TNM needed more than a generic conversational interface. Leelou AI had to reflect its coaching posture and improve through testing with people who understood its quality standards.