TiCON brings first Hardware integrated multilingual chatbot (Bengali, English, Korean, Japanese and it can be extended to any other languages), we intent to utilize full power of Big Data, NLP, ML & Speech Recognition technology.
Machine Learning (ML)
Natural Language Proccessing(NLP)
Our Own Algorithm
Speach Recognition(STT & TTS)
TiCON brings first Hardware integrated chatbot, in Bengali, we intent to utilize full power of Big Data, NLP, ML & Speech Recognition technology.
We can now schedule alarm, schedule/reminder, ask our bot to search anything for us, call to any contact or any number, Open any other app etc.
When making small talk with THEIA, bot can communicate with you with some real nice response. It can be enjoyable at the same time it can be very informative for you.
THEIA is multilingual. Currently it works on Bengali, English, Japanese & Korean. And have the ability to expand to any other language.
THEIA is first bot integrated with hardware based 3D model in Bangladesh & one of the few in the world. We are using Android box for hosting the 3D model.
We recently started taking public input while making a conversation, as our plan to work with Big Data & NLP. The more data we have, the better THEIA will be.
The Process We Folllow
- Utilize Apache SparkML
- Use JAVA wrapper for Android
- Initial data inventory for Machine Learning
The AI Part
- Grow data inventory based on conversation & user input
- Unsupervised model for auto learning from input
App (Android & iOS)
- Android app for frontend functionality/user engagement with ML
- Android connects with Python/Django backend
User calls to our contact center
We record all conversation for our future ML/AI implementation
We convert all voice conversation to text
Then we save our all conversation to our database
We implement ML to all recorded conversation, for pattern understanding
Then we implement AI to auto answer user queries using our ML technology
We can deliver all support question answers automatically, given there is enough data & pattern
Text to speech implementation can also be done following same protocol