AI-Powered Memorial Companion Application Development

Develop an AI Memorial Companion App with Real-Time Animated Face, Voice Cloning, AI Conversation & Memory

Project Description

I am looking for an experienced AI/mobile app development team or highly skilled developer to build the first version of Aevomi, an AI-powered memorial companion application.

The purpose of the app is to allow users to preserve memories of loved ones who have passed away and interact with an AI-generated animated digital representation of that person.

This is not intended to be a traditional chatbot. The main experience should feel closer to having a face-to-face video conversation with a personalized AI representation of the loved one.

Core User Experience

A user should be able to create a memorial profile for a deceased family member, friend, mentor, or other loved one and upload information such as:

Photographs
Short voice/audio recordings
Videos
Written messages, letters, emails, or text conversations
Stories and memories
Important life events
Personality traits
Values and beliefs
Hobbies and interests
Common phrases
Nicknames
Catchphrases
Typical advice
Communication style
Sense of humor
Family relationships

The system would use this information to create a personalized AI memorial companion.

Most Important Feature: Real-Time Animated Person

The animated face must be part of the initial product, not something planned only for a future version.

Using uploaded photographs, the app should create a realistic digital representation of the person.

When the AI speaks, the face should animate naturally, including features such as:

Realistic lip synchronization
Eye movement and blinking
Head movement
Facial expressions
Smiling
Natural listening behavior
Appropriate emotional expressions
Natural pauses while listening and speaking

The user should be able to see the animated person on the screen and speak to them using the phone's microphone.

AI Voice Recreation

Users should be able to upload one or more voice recordings of the deceased person.

The system should use AI voice-cloning or voice-generation technology to generate speech that resembles the person's voice.

The animated face and generated voice must work together so that the user both sees and hears the digital person speaking.

Conversational AI

The application should integrate an LLM or similar conversational AI system.

For example, a user might say:

"Grandpa, I had a difficult day today."

The AI should respond based on the person's uploaded memories, personality, values, communication style, expressions, and typical advice.

Users should be able to have natural voice-to-voice conversations without typing.

A basic interaction could be:

User speaks → Speech recognition → AI understands the conversation → Relevant memories/personality information retrieved → AI generates response → Voice generated → Animated face speaks response

The response time should be fast enough to make the interaction feel natural.

Personality and Memory System

Each memorial profile should have a structured knowledge base containing information about the person's:

Personality
Memories
Values
Beliefs
Habits
Interests
Expressions
Relationships
Stories
Humor
Communication style
Typical advice

The AI should retrieve relevant information from this memory library when responding.

For example, if the user asks:

"Do you remember our trip to Italy?"

the system should be able to retrieve previously uploaded information, stories, photographs, or memories related to that trip and use them as conversation context.

Long-Term Conversation Memory

The application should also be capable of remembering previous conversations, with user permission.

For example, if the user tells the companion:

"I have a job interview next Monday."

during a later conversation, the companion could ask:

"How did your interview go?"

This long-term memory is important because conversations should have continuity rather than starting from zero every time.

Family Contributions

Eventually, multiple approved family members should be able to contribute memories, photos, videos, stories, and other information to the same memorial profile.

Different family members may have their own private conversations while using the same shared memorial knowledge base.

Please explain whether you recommend including this capability in the MVP or adding it immediately after the first release.

Privacy, Security & Ethical Requirements

This application will contain extremely personal information, so privacy and security are critical.

The system should include:

Secure user authentication
Secure cloud storage
Encryption of sensitive information
Protection of photos, videos, voice recordings, memories, and conversations
User-controlled data
Ability to permanently delete a memorial profile and associated data
Appropriate permissions and access controls

The application must also clearly communicate that the animated person is an AI-generated memorial representation, not the actual deceased individual.

The AI should not claim that it is conscious, communicating from an afterlife, or possesses information beyond what was provided to the system.

The system should also avoid presenting invented memories as factual memories.

Appropriate safeguards should be considered for medical, legal, financial, and other high-stakes advice.

Technical Components

I expect the application to require integration of technologies such as:

iOS/Android mobile application
Backend/API development
Large Language Model integration
Retrieval-Augmented Generation (RAG) or similar memory retrieval system
Vector database / memory storage
Speech-to-text
Text-to-speech
AI voice cloning
Real-time facial animation
Lip synchronization
Photo-to-avatar/digital-human generation
Secure cloud storage
User authentication
Conversation history
Long-term AI memory

I am open to using reliable third-party AI APIs/services where appropriate rather than developing every AI component from scratch.

Initial Deliverables

For the first phase, I would like the freelancer/team to:

Review the complete concept and recommend the best technical architecture.
Recommend appropriate existing AI APIs/services for the animated face, voice cloning, LLM, speech recognition, and memory components.
Design the UI/UX.
Develop the backend infrastructure.
Develop the mobile application.
Implement memorial-profile creation.
Implement photo, audio, video, text, and memory uploads.
Implement AI personality and memory retrieval.
Implement real-time voice conversations.
Implement AI voice recreation.
Implement the real-time animated digital face.
Synchronize the animated face with AI-generated speech.
Implement conversation memory.
Implement privacy, security, and data-deletion controls.
Test and optimize the complete user experience.
Provide source code and technical documentation.
What I Need From You

When submitting your proposal, please explain:

Your experience with AI-powered mobile applications
Your experience with LLM APIs
Your experience with RAG/vector databases and AI memory
Your experience with voice cloning or AI speech generation
Your experience with digital humans, talking avatars, facial animation, or lip synchronization
Examples of similar projects you have completed
The technology stack you recommend
Which third-party AI services/APIs you recommend
Your proposed development phases
Estimated timeline
Estimated cost
Estimated ongoing API/cloud costs
Whether you are an individual developer or a development team

Please do not submit a generic proposal.

In your response, specifically explain how you would technically create a realistic animated face from uploaded photographs and make it speak in real time using an AI-generated version of the person's voice.

This component is essential to the project.

The goal is to create a polished, scalable foundation for Aevomi that can eventually grow into a private digital family legacy platform, preserving photographs, voices, videos, stories, memories, and personalities across generations.