loder

CEO Prosa.ai interviewed by Cybernews: “Speech recognition could be used by every person”

Language barriers might soon become a thing of the past with the development of advanced speech recognition tools.

Chatbots and artificial intelligence are constantly learning, erasing not only geographical but also language barriers between individuals. Whether you are a professional IT specialist, a bank worker, or a graduate student, there are many products available to you if you’re keen on trying AI-governed solutions for speech recognition.

To talk about the importance and development of speech recognition and natural language processing, Cybernews contacted Teguh Eko Budiarto, CEO of Prosa.ai – an AI-incorporated solution which develops natural language processing with deep learning that allows computers to process information in Indonesian.

Let’s go back to the very beginning of Prosa. Can you tell us more about your journey since 2018?

Prosa.ai was established based on research by dedicated founders, Dr. Ayu Purwarianti and Mrs. Dessi Puji Lestari Ph.D. They have been interested in researching Natural Language Processing (NLP) and Machine Learning for more than 20 years since they studied in Bandung Institute Technology and sharpened their research at reputable technology universities in Japan to get Master and Doctoral/Ph.D. degrees.

Dessi and Ayu expressed and convinced me that although AI experts are still limited in Indonesia, they believe Indonesia can develop AI technology to accelerate the local economy by enhancing business services and bringing better value to customers. They emphasized that making it happen requires collaboration between academia, industry, and all stakeholders.

In 2018, we solidified our steps and agreed to establish a startup company in the field of artificial intelligence with a specialization in Bahasa Indonesia. We named the company Prosa, which is the abbreviation from “Pemrosesan Bahasa” (Indonesian Language Processing).

Can you introduce us to what you do? How is AI incorporated into your solutions?

Firstly, with the technology that we have, we have been developing a Natural language processing (NLP) toolkit API for businesses in Indonesia that consists of conversational AI, speech processing, voice biometrics, and etc. This NLP Toolkit Solutions would be useful and add value for businesses, as it is easy to integrate with various applications.

Secondly, because we realize most Indonesian businesses aren’t ready to use the toolkit themselves, we have been developing end solutions text and speech processing software capabilities and voice biometrics for companies such as startups, SMEs, corporations, banking, Insurance, and the government.

Thirdly, we also have been developing solutions for end clients who need efficiency and simplicity for regular activities or works such as platform text to speech to make voices over that are popularly used by many content creators. Besides that, we are also developing a platform for speech-to-text, we name it Salyns, an automatic transcription that can be used by anyone such as a student, worker, content creator, journalist, etc.

In your opinion, which industries could greatly benefit from implementing speech recognition solutions?

Speech recognition could be used by every person. Speech to Text (STT) and Text to Speech (TTS) are products of voice processing that are used by creative industries, mass media, etc. Another Product is voice biometrics which is currently used for customer data security, such as BFSI (Bank, Financial, Service, and Insurance) started to look at voice biometrics as an important complement to other biometrics to recognize customers. It is usually done in the customer onboarding process.

In various industries or manufacturers, offices, and governments, voice biometrics are used to identify the employee in attendance systems, market analysis, sentiment analysis as well as fraud detection for educational systems such as in exam certification, etc. Speech recognition along with other biometrics could be holistically used for end customers such as for the verification of residents or guests for smart door locks in the smart home systems.

How do you think the present global events affected your field of work?

Just two years after Prosa.ai was established, in 2020, coronavirus was declared a global pandemic and affected our business. Many orders, especially from the government, were canceled because they focused on pandemics and mitigation. The condition that suddenly digitalizes on the other side changes our business direction, not only focusing on B to B but also focusing on B to C.

The Global Pandemic effect on the reduced physical interaction makes some people use digital tools to communicate with others. AI technology is getting closer to being used by humans. For this reason, we are accelerating the development of Text to Speech and Speech to Text to help people in their daily life as well as enhance NLP development for businesses and industries to be more understanding of customers who use various tools or media.

Besides natural language processing solutions, what other technologies can significantly enhance business operations?

Marketing methods that shift from conventional practices make digital marketing the best strategy to enhance business. Besides that, to increase customer satisfaction, customer retention, and business operation efficiency, CRM software integrated with an omnichannel chatbot is necessary for businesses.

Talking about cybersecurity, what would you consider the key best practices every company should follow?

We got ISO 27001 for IT Security Management System in March 2022, but we commit to data security since Prosa.ai was established. Before running the project, we always make an agreement of confidentiality with partners.

We also practice data security, such as obligating all employees to use PCs that we give, practicing BYOD (Bring Your Own Device), cloud data security, zero-trust security, etc. The important thing, we should commit to comply with ISO 27001.

What challenges do you notice companies run into on their digital transformation journey?

For an AI company, the limited talent of AI is still the biggest challenge. One strategy to attract young talent to understand AI and its operations is an apprentice program. Many of our employees start their careers in our company as an apprentice and many of them have positions as managers under 30s.

Second, in this transition of digital transformations, AI companies must have extra patience to explain to various consumers how AI works. Many end consumers still think that AI technology is humans with high intelligence and understands anything without collecting data or training by customers as the user.

In your opinion, where can we expect to see AI-powered solutions be used more often in the near future?

In line with digital transformations, human services are starting to be replaced with AI, especially BFSI, and E-commerce is starting to use robocalls for Interactive Voice Response (IVR) and using holistic biometrics for consumer data security.

Possibly after that, AI technology will be used by humans for daily life, such as waking up by directly using a virtual assistant to remind the daily schedule or hearing news from a virtual anchor, etc.

Would you like to share what’s next for Prosa?

We are aligned with the accelerated digital transformation that the government is promoting, which is leading to a shift from conventional to optimized services with AI implementations.

Hopefully, Prosa can make a real contribution towards a glorious Indonesia in 2045, as initiated by our government.

This article is taken from Cybernews (permission has been obtained to distribute it) so it can be accessed again via this link

None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None