For 3 years, Akshaya Aron and Dipankar Sarkar worked on herbal language processing (NLP), Machine Learning (ML), neural networks and Artificial Intelligence (AI). This helped them in making the first completely automatic bot on WhatsApp in which you will upload a variety of, and begin talking to it. They ran it for 4 months – got 40k customers without spending any cash, were given five million messages exchanged at the platform – and ultimately close it down, for no longer being a feasible enterprise.
But, this is not the stop of the story.
Earlier in April this yr, the duo received tempo again and got here up with Octo.Ai – to begin with as a platform for absolutely automating a enterprise’s internet advertising and marketing wishes (push notifications, personalised feeds, etc.) – and later in July as an analytics platform constructed for Machine Learning.
They also currently raised Seed funding of $200K at a valuation of $1.Five Mn. The funding became led by means of a number of angel buyers, which covered Rohan and Arjun Malhotra (Investopad); Rahul Khanna (Trifecta Capital); Rakesh Agrawal (US based totally- seed in Lyft, Cruise, Shyp, Poshmark and Lendup), Sidharth Rao (WebChutney), Outbox Ventures, Rajan Navani; Gautam Gandhi (Ex-Google, Angel Investor); Jaspreet Bindra; and Gagan Duggal.
The team of seven is presently based totally in New Delhi.
Journey From A WhatsApp Bot To An Open Source Analytics Platform
The duo to start with decided to use personalisation algorithms and rationale seize algorithms they had made for the WhatsApp bot, and launch Octo as an automatic advertising and marketing answer. “We have been presenting intelligent push notifications and personalized feeds at the side of the same old dashboard, segmentation, and so on. However, our initial sales cycle, made it apparent that some of the assumptions we made had been wrong,” stated Akshaya.
The one critical assumption was that the records would be to be had to run algorithms. As Akshaya explains it further, excessive-quit person personalisation problems require records to be based efficiently. But while a corporation begins, the heavy dependence on loose/ freemium 1/3-celebration analytic offerings – Google analytics, Mixpanel, and others – outcomes in heavy information loss as most analytic vendors lock it into their structures.
In this scenario, they could need to be within the employer’s stack for 3-6 months. They figured even as selling that every patron ( employer customers mainly ) don’t need to proportion their facts with startups and the brink for entry is excessive.
“At the start of this month, we open-sourced our product code, it is a combined model of MixPanel and Segment.Io. Any startup/employer can installation it on their very own servers and personalize it as properly. This has been carried out to provide control of facts back to the agencies and enable everybody to apply ML & AI on their facts sets. Till now they have got all been locked in on paid 1/3 party services” stated Akshaya.
Here is the way it works: (before the usage of Octo)
Growth Hacking Companies’ Marketing Strategies
According to Dipankar, the framework of Octo collects all the analytics information and permits actual records technology/ML going ahead. Also, as one length doesn’t fit all, businesses need freedom to adjust records structures to deal with algorithms better ideal for their business case.
“We are open-supply, subsequently infinitely customisable, as in keeping with the need of the client. This lets in groups to increase the machine for his or her own inner Machine getting to know algorithms and workflows. We estimate over 70% discount in their external SaaS costs after they start scaling,” he brought.
Majorly, companies could make the satisfactory use of Octo in four possible ways:
Use it as their in-residence analytics answer: Instead of the use of Mixpanel, GA or any other third-celebration equipment they can just install Octo, placed the only line code on their website and apps and feature their very own version going for walks! The largest gain is proudly owning the records and also being capable of customize the reviews as in step with enterprise want.
Autonomous push advertising/advertising and marketing automation: Companies can install Octo, and ship personalised push messages based totally on triggers to each person to force more energetic customers and engagement. Like all advertising automation platforms, Octo comes constructed in with a complicated rule engine and a complicated routing mechanism to deal with all advertising communication. Eg. If Octo’s analytics show a person is attractive much less, an automated notification can go to them to get them returned on the machine.
In-app and on-web site personalisation: Companies can use the suggestions and phase-of-one functions to personalise their app and internet site for each user, just like Facebook or Netflix. Octo increases time on web site and revenue by using imparting content material from the machine a good way to engage the person.
Marketplace dealer analytics: Marketplaces can use it to strength their vendor analytics thru a multi-tenant and fantastically scalable machine. The gadget can also offer dealers with suggestions for merchandise and key insights on how their products are doing.
Monetisation, Competition And Challenges Ahead
The sales model of Octo is just like other open-source agencies, like MongoDB, and Red Hat. They will rate agencies for customisations (traditionally enterprise clients who want a custom answer built for themselves + technical aid), hosted answers in their open-source supplying, consulting and education on Octo.
The group is currently competing with Adobe Omniture, Google Analytics, Mixpanel, Segment.Io, Kahuna and numerous different groups.
When requested about the challenges, Akshaya responded, “The biggest assignment with any open-source era is adoption. As we're a brand new organization, massive organizations who could benefit a lot from such era are a touch hesitant in adopting us. Other than that the lengthy choice cycles in bigger businesses is always something which hurts small startups.”
Pushing Internet Marketing To Beyond Basic Software Stack
The team currently has an inner goal to reach one hundred customers by means of the end of 2016.
The Octo founders aim to push Internet advertising to the subsequent level a good deal past the fundamental software stack, in which machines are autonomously taking actions based totally on person behaviour insights. Right now some of these tools are utilized by advertising groups to make inferences and carry out actions. The duo similarly believes that an AI gadget should perform those movements because it will be extra green and obviously extra personalised.
“In India, we are one of these few crew startups who're certainly going to do essential adjustments. We need Octo for use as Mangodb, or SQL, so that whenever any developer builds an app, they generally add an entire code library to it, and open source definitely helps in it,” said Dipankar.
According to a June 2016 document by way of Nasscom, the Indian analytics industry is predicted to reach $16 Bn with the aid of 2025, from $2 Bn currently. India is currently some of the top 10 destinations for analytics, with AI and deep mastering algorithms creeping gradually in all factors of even the pleasant of the gang.
The coup de grace in it is section-of-one advertising and analytics-driven advertising automation. Industry veterans consider this market to be headed in the direction of use of Machine Learning and neural networks on information sets. This will assist corporations to serve customers better and to improve their enjoy in a cellular application or on a internet site – with startups like Octo round to facilitate the technique.