Top-11 Artificial Intelligence Startups in Indonesia
For a country with 17,508 islands, it’s quite surprising to think that only three other countries in the world have more people than Indonesia does. It’s a country with a rich history, a growing population, and four unicorns grazing its lush island pastures. That’s right, of the world’s 311 unicorns (that number is accurate according to CB Insights as of today), four of them have roots in Indonesia. Late last year, we sent one of our MBAs over to spend a few weeks in the capital city, Jakarta, looking at the Indonesian tech scene. That research spawned an article on GO-JEK, a simply fascinating company. It also led to lots of research around how Indonesia is competing in the global artificial intelligence (AI) race.
At first, we were thinking about naming this article “all the AI startups we could find in Indonesia,” but then we’d get dozens of emails for the rest of the year about all the hidden gems we “missed.” Instead, we sat down and did some Crunchbase searches, combed through company websites, did some asking around, talked to some of the local startup founders, and as a result, we have below what is our best estimation of the top AI startups in Indonesia today. If you are one of the below startups, feel free to celebrate your acceptance to this top-11 list by emailing this article to every single person you know.
|Sonar Platform||Social Media Monitoring||Jakarta||.15|
|Dattabot||General Big Data||Jakarta||N/A|
|Eureka.ai||Telcom Big Data||Jakarta||N/A|
|AiSensum||Robotic Process Automation||Jakarta||N/A|
The above startups should be proud of what they’ve accomplished because each of them stood out, in some way, against the total number of companies our foreign correspondent pored over while relaxing in some of North Jakarta’s finest health spas. Let’s take a closer look at each of these startups.
If the name Snapcart rings a bell, it could be because you read about them in our article last month on Smart Receipts and Why We Should Use Them. Founded in 2015, Indonesian startup Snapcart has taken in $14.7 million in funding so far to create a mobile application that gives shoppers cashback for scanning their receipts. This allows the company to collect massive amounts of purchase data, then analyze it and offer real-time insights to big names like Johnson & Johnson, Unilever, P&G, and Nestle. Snapcart currently operates in Indonesia, Philippines, Singapore, and Brazil. (Sounds like something GO-JEK might be interested in getting their hands on.) With high retention and engagement rates, Snapcart is also able to send targeted surveys to customers asking them relevant questions at the right time.
The system can also capture transactions from independent chains where existing solutions do not capture, and to-date they’ve processed over a half a billion receipts.
Founded in 2015, Jakarta startup Kata.ai has taken in $3.5 million to build Indonesia’s number one conversational AI platform. A case study they published talks about the success Unilever had when deploying a chatbot to engage with customers. The female chatbot persona was named Jemma, and was deployed on Line messenger, one of Indonesia’s most popular messaging apps. Less than a year after its deployment, Jemma managed to acquire 1.5 million friends, with more than 50 million incoming messages in 17 million sessions. “Some of them even tried to confide their dreams and problems to her,” said the case study, and the longest conversation recorded exceeded four hours.
Another case study discusses a chatbot deployment by Telkomsel, Indonesia’s largest cellular operator with more than 120 million subscribers (that’s almost half of Indonesia’s population). Turns out 96% of customer inquiries can actually be handled by the chatbot with minimal human interaction. In order to scale more quickly, the company built a very slick platform that makes it easy for anyone to build a bot.
We talked with Kata.ai’s CEO and Co-Founder, Irzan Raditya, about why conversational AI is so popular in Indonesia. He said it’s largely because the big tech players are behind the game when it comes to Natural Language Processing (NLP) for Bahasa Indonesia (that’s the language they speak in most of Indonesia). It’s not an easy task when you’re trying to understand a language that has 13 different ways to say “I.” When companies like Accenture partner up with a “small” firm like Kata.ai to bid on projects, it helps demonstrate that they’re best-of-breed.
Moving on to our second conversational AI startup that speaks Bahasa Indonesia, we have BJtech. Founded in 2015, the company has taken in $1.5 million in funding so far to develop an easy-to-use platform that helps you create chatbots for your business. Their first product is a virtual friend that does things for you and expects nothing in return, and an intelligent banking app. Clients include Uber, Skyscanner, and Zomato, though we have no idea what Uber is doing speaking the Indonesian language after GO-JEK showed them the door. There’s a fair amount of Engrish on their website, so they may want to sort that out because that’s not the best look for a language processing company.
Founded in 2015, Sonar Platform has taken in just $150,000 in funding to develop a social media monitoring platform that – you guessed it – speaks Bahasa Indonesia. As an example, Unilever Indonesia certainly doesn’t want some loudmouth influencer bad-mouthing their latest skin-whitening product, and in order to see what people are saying about their products, they might use a platform like this one. The platform allows you to monitor social media in real-time, and they process over 1 million conversations a day, all of which can be mined later for insights. Their platform can gauge sentiment as well, and Air Asia uses it to monitor how pissed off people get when their flights are delayed.
Moving away from the Bahasa Indonesia theme for a moment, we have a startup called Nodeflux that was founded in 2016 with an undisclosed amount of funding which they’re using to develop Indonesia’s first intelligent video analytics platform. Backed by Telkom Indonesia, they’ve also partnered with NVIDIA to offer video analytics services to companies like GO-JEK which uses their service to monitor CCTV cameras on the streets of Jakarta to track where the 1 million-plus fleet of GO-JEK scooters is at during any given time.
They also offer services like facial recognition, license plate reading, flood monitoring, and trash detection.
And we’re back, on to more conversational AI for Bahasa Indonesia with the aptly named Bahasa.ai, a startup that was founded in 2017 and which has taken in an undisclosed amount of funding to “build the most robust NLP modules for Bahasa Indonesia.” Based on the AI research focus we observed at Kata.ai, they have their work cut out for them. Since our own Bahasa skills are lacking, and they haven’t translated their website (can’t they get some of their algos to do it?), that’s about all we can tell you about Bahasa.ia. Oh, and one of their competitors vouched for their capabilities which was awful nice of them. In other words, they’re not just a company that creates chatbot scripts and says they use AI when they actually don’t. (We’re told there are some of those out there in Jakarta but we’re not naming names.)
Our next company we know little about because they’re so new. Founded in 2018, Prosa.ai was founded by Indonesian experts in AI for NLP in text and speech. They already have subscription pricing on their website, so we can only assume that they have developed a product. We saw that they’re backed by a notable Indonesian venture capitalist, so we can also assume that someone vetted their business model against the plethora of NLP startups that are already tackling this problem.
Founded in 2003, Indonesian startup Dattabot – formerly known as Mediatrac- is big data analytics company with an undisclosed amount of funding that has assembled the most comprehensive data library in Indonesia. We sat down with the founders, Regi Wahyu and Imron Zuhri, who told us how they started out scanning Indonesia’s dark world of data, largely offline and in printed form. In 2010, they began scaling their data offering and in 2015, pivoted to become the company they are today that targets a number of industry verticals.
Their first project involved a large FMCG company with three databases of data and no desire to spend money on building a data warehouse. Dattabot used some clever AI algorithms to solve that problem, and revenues soared as they optimized various aspects of the operation like the “traveling salesman problem” we discussed before. Then came one of Indonesia’s largest telcom providers with a big problem. More than 90% of accounts were prepaid. How can you know the customer? Dattabot used AI to solve that problem too. That’s when they realized that an even bigger opportunity could be found in Indonesian farming, an industry that consists of 49 million farmers that represent 41% of the country’s total labor force. Their subsidiary Hara.ag was then born, and the story behind it is so interesting we’re going to dedicate an entire article to it. Stay tuned.
We actually don’t know when our next company was founded, or how much funding they’ve taken in, but we do know their PR company is asleep at the wheel because they never responded to our email asking for more info. That’s okay though, because when you’re busy kicking a33 and taking names, who needs PR anyways? The man at the helm is Benjamin Soemartopo, previously with McKinsey & Company for 12 years as Managing Partner and CEO for Indonesia and before that, Managing Director for Standard Chartered’s Bank Private Equity in Indonesia for six years. The company enables partnerships between mobile operators and companies in industries including banking, insurance, transportation, and consumer goods with a global presence:
That’s the who/what/where, and about all we can tell you for now.
Update 02/18/20: Eureka AI has raised $20 million in Series B funding to support international expansion into Europe and the U.S., as well as further develop its product portfolio. This brings the company’s total disclosed funding to $20 million to date.
Our second to the last startup was somewhat difficult to understand until the company emailed us to clear things up. Their main source of revenue is data monetization partnerships through their platform called Octopi, a machine learning-driven SaaS dashboard that creates business intelligence insights. The firm also offers Robotic Process Automation (RPA) that they describe as “low cost bets for companies who are unwilling or unable to invest in fully automated AI platforms.” They also let us know that they didn’t appreciate us making fun of their octopus, something we blamed on our ethnocentric tendencies to make fun of things we don’t understand – like this diagram.
Joking aside, they’re enthusiastic about what they’re doing so we may go visit them when we’re back in Jakarta. They also have a sister company called Neurosensum which uses AI for consumer research and which may have some toys we can play with.
Last but not least is a startup called Deligence.ai. We know almost nothing about them because they’ve been so busy doing AI stuff that they haven’t even created a profile on Crunchbase. The only reason they made this top-11 list is because a founder we talked to vouched for them. (See how important networking is kids?) According to the website, they provide “organizations the most optimal access to the cutting-edge computer vision, machine learning, and big data technology.” We’ve also reached our word limit on this article so time for a conclusion.
Forgetting about AI for a minute, we were simply floored by the opportunity that we saw in the world’s fourth-largest country, the talented and passionate people we spoke to who could see the opportunity, the astounding success of startups like GO-JEK, and conversely, how isolated and relatively untapped the tech scene seemed. (We’re trying desperately to find emerging technology startups of any kind in the country’s second-largest city, Surabaya, and have come up empty-handed so far.) In the future, we’re going to take a closer look at what sort of investment opportunities might exist for retail investors in Indonesia – largely in the area of ETFs – and also deep-dive into the fascinating world of Indonesia’s “big” data problem and how it’s being solved.
Despite what the pundits say, FAANG stocks (Facebook, Apple, Amazon, Netflix, Google) don't give you real exposure to AI. Read about 7 stocks that give you true pure-play exposure to AI in our guide to investing in AI healthcare companies, freely available to Nanalyze subscribers.