mScriber is a Delhi based startup with very strong expertise in Speech Recognition Technologies, Processes and Applications. 
I have profiled Ubona, a voice recognition system earlier – but mScriber’s USP lies in the several other things that they do after the voice processing:
It listens carefully to what you have to say.It then applies speech recognition technologies and processes to convert your voice into text. Finally, it lets that text reach your recipient(s) in the way you want – as an SMS, Email, and Blog. It also lets your recipient(s) use the text the way you want – for general information purposes, for searching for additional information, for voice analysis or as simple notes.
The core of their offering is not just the technology, but solutions built around that, and to me, that’s a killer!
The opportunities for mScriber are huge – right from tying up with telco operators for Rural VAS (Rural IVR is the next big bet for Indian telecom operators) to local search providers, to any damn service that wants to tap into voice, mashup with their content and deliver to any other medium.
mScriber is targeting emerging markets and currently offers following solutions:
- Voice2SMS : send your SMS without the need to type it, from any phone
- Voicemail2SMS : a call completion product that converts a landline or mobile phone voicemail in to text and sends it across as an SMS,
- Voice2Search : converts voice prompts into text and completes the transaction (Alabot is in this space as well)
- Voice2Notes : narrate your notes by calling up mScriber Voice2NotesT telephone number, leave an email address where those notes must be emailed. mScriber Voice2NotesT coverts your verbal notes into text and sends it to the email address you left.
- Voice2Analytics: analyse the voice data by first converting it into text using mScriber speech recognition technologies and processes.
mScriber has received investment from Ace Global Services and Perfect Future; and has a good mix of talent in founding team.
This is one of the startups to watch out for.
What’s your take?
looks interesting …tried their Voice to SMS service . good attempt i must say . In first attempt i said something but forget to say my cell number [i was calling from my cell only] so i got a SMS on my cell , transcript was fine .
In second one i said the same message but appended my cell number in the end . transcript was right but my cell number was added in that incoming message .
third time i spoke in hindi [even my number] I got an SMS but without a message .
Fourth and last time i called from a Land line spoke in Hindi -english mix said my cell number in English .got a message. but my number was again included . but there were no details of sending party .hindi part of my message was removed
idea surely have a lot of potential specially the third use case . I can see this being a way to tell your “Dill Ke Baat ” to the Girl Next Door without getting traced
.
good luck to mScriber Team !!
yeah: “I can see this being a way to tell your “Dill Ke Baat ” to the Girl Next Door without getting traced” will be used for about 30% sooner by youths.
This is Great product company which i like along with Ubono.
interesting to see Speech Technology review here, will put to my blog “Business @ Speed of Voice” http://www.raxitsheth.blogspot.com
I am in office and right now (9:30 pm IST)
i try to call this number, but the number is operated on Weekdays till 6:30 pm only. [may be bcoz Its in alpha] So right now i am unable to check,[And i hope they are not doing "Manual Speech recognition" 
kidding ... disclaimar: i know few folks from mscriber.]
Natural speech recognition (without grammer/option) is very tough to recognize, including some “who is who” like MSFT,Nuance, IBM and more doing research, if it works, Wow ! Especially very less Speech recognition for Indian languages, if it also work for indian languages its Big Wow !
Congrats to the team !
Cheers,
Raxit
Great work by the team. The really killer thing will be if they have the databases and grammar’s for Indian Languages and accents. There’s a huge demand and need for that.
I think they are just collecting the speech database to train their system. This is the standard way of doing so. Even Kirusa does it the same. For varied or mixed languages, the acoustics is the more important than the grammar. Good start.
Pingback: Ubona secures Series A Funding from Capital18 |Technology and Business Startups in India
too late to give a review on this but I’m simply amazed by this.
there are, and i was expecting them, a few issues with the service, foremost being the process-flow understanding. as Prashant said, it displays the mobile number too (which is a bit wierd) plus the obvious learning-phase errors as the scriber learns new words/accents/languages.
terrific service btw and am sure future’s looking good for such innovation. it can only get better from here (no, i don’t mean what i didn’t mean)
Another speech recognition company to watch out for is Simmortel Voice . As for the others, clearly they are still in the stage of collecting acoustic data. Is Kirusa working for the Indian market, btw?
Kirusa is powering VoiceSMS all over india.