It appears, Angam Parashar, Ankit Narayan Singh and Muktabh Srivastava, did not leave any stone unturned to prove this adage.
When the trio began their entrepreneurial stint in January 2014, they had no idea what the future held for them. What started as only a simple idea to resurface vintage content in a contextual way for publishers, later grew to become to a dive into the complexities of deep studying and AI.
Now, after years of starting their Gurugram-based totally venture ParallelDots, has been featured along pinnacle deep learning startups by using Denny Britz in his blog. The enterprise was also adjudged as the pinnacle three startups in India at the Seedstars 2014 occasion. Not most effective this, Muktabh is one of the maximum regarded authors on topics like deep studying and synthetic intelligence on Quora globally and became additionally one of the audio system at The Fifth Elephant convention – 2015, India’s largest large-facts and machine getting to know convention, held at Bengaluru.
The Beginning: From A Simple Tech Startup To A Deep Learning Platform
Before starting ParallelDots, Angam become operating as a consultant at Opera Solutions, a boutique huge-information consulting company and Ankit changed into working in Australia. Being near batch buddies or even wing associates at IIT Kharagpur, their friendship had withstood the check of time and distance – a reality that is pondered when Ankit left his soft process, got here again to India to aid Angam together with his endeavour.
While they were brainstorming on initial thoughts, they also began searching out a tech expert to sign up for them. Their search ended with Muktabh, who changed into Angam’s colleague at Opera Solutions. A BITS Pilani graduate, he was a records scientist there and eventually joined ParallelDots because the third co-founder.
The journey began with ParallelDot’s selection in T-Labs’ fifth batch in January 2014. During the inception, they labored with on line publishers to create timelines for any occasion the use of their archived content material. For instance, if it were a information internet site that became overlaying the ‘US Elections 2016,’ they may fast accumulate all critical content material from the past in a timeline to present a fully immersive, complete backstory.
They started out with this idea and went on to three pivots afterward, but, they quickly realised that to create accurate timelines they would need a absolutely powerful technology. That was the time they began operating on deep gaining knowledge of. They effectively constructed the technology that could create the timelines accurately. However, as they talked to increasingly humans, they realised the need of such era in other domain names, specially inside businesses.
Angam in addition remembers,
“This become 2014, while most people didn’t even know what deep getting to know was. So, we were one of the earliest groups who certainly started out the use of it.”
Since then, the ParallelDots trio have come a protracted way. They have advanced quite a few generation based on the remarks they accrued from the market. They have additionally modified their enterprise models more than one times as an test, to figure out the right element.
“However, we are nevertheless in stealth mode – the manner I see it's miles that we still tweak our business model simply to see what is running higher and what isn’t,” said Angam.
Core Offerings: Media Monitoring Tools, Deep Learning API’s For Text, And More
Currently working with a team of 25, the employer offers a complete stack of wise text analytics APIs to companies. These APIs can now offer one-of-a-kind use cases relying upon the want of the business enterprise or writer.
One use case is inside the shape of a advice engine. As defined through Angam, “Consider those two sentences – ‘Sachin is the greatest batsman’ and the opposite, ‘Tendulkar is the finest cricketer.’ One can see that here no phrases are commonplace but nonetheless they're the same sentence. Now imagine, if we will build a search engine on pinnacle of this generation and use it to advise contextual content material, ads, and many others. The consequences are impartial of the words used, impartial of the tags – they are purely contextual.”
Also Read:- Do You Make Any of These 10 Simple SEO Mistakes?
Next, some customers use their technology to extract entities which might be present in the article to understand what it talks approximately, even as some use it for extremely correct sentiment evaluation generation. “Say there are sentences – ‘I even have a weigh down on Virat Kohli’ and ‘Virat Kohli overwhelmed my hopes’. Now, the latter is a totally bad sentence however the verb that is figuring out the sentiment occurs to be the identical – crush. Our set of rules understands the that means of the sentence – it may crack all this quite effortlessly,” simplified Angam.
Furthermore, the usage of all their deep Learning APIs, they have built a media tracking tool too. As stated through Angam, “We move slowly thousands and thousands of resources across the net, Twitter, Facebook, and so on. To monitor unique occasions, key phrases, hashtags, and so forth. We additionally have built an anomaly detection algorithm that mechanically increases indicators each time there’s an anomaly inside the occasion that our client is tracking.”
The group also gives custom answers to businesses who need to educate deep mastering fashions on their proprietary dataset. Angam noted, “We have a completely strong presence on applicable communities like Quora, Reddit, Hacker News, and so forth. This facilitates us pull organizations across the globe who are trying to utilize the energy of system studying on their current statistics”.
“It’s Hard For AWS, IBM Softlayer, To Build Plug And Play Kind Smart APIs”
In the outside world, from a era standpoint, systems like IBM Softlayer, AWS, and so forth. Are themselves looking to build such clever APIs within their cloud structures.
However, as said by means of Angam, it is difficult for them to build clever APIs that may be used as a plug-and-play device. Every business enterprise has its own dataset which is exclusive than the others. Given the truth that each one system getting to know models work quality on the dataset that they're trained, pre-current models can't supply the nice effects on one of a kind datasets, unless the fashions are explicitly educated on them. Moreover, organization shipping often requires numerous customisation and continuous guide and the call for of such answers within corporations is developing globally.
He further brought,
“The price proposition is straightforward – we provide better technology without any development effort therefore saving time and money for our customers. There are some present media tracking solutions that are used by enterprises inclusive of Brandwatch, Keyhole, etc. Which provide gear to music any occasion/keyword throughout the internet. However, our proprietary algorithms have allowed us to show insights which aren't to be had within current answers. Technology like sentiment analysis, motive analysis, contextual classifiers, troll detection are a number of the functions we've constructed for our product.”
When requested about the challenges being confronted inside the preliminary phase, Angam said expertise crunch changed into the high one. “This entire area is very new. It’s difficult to discover proper talent at the price that we can manage. We overcame this venture by means of supplying the exceptional quality paintings to our team which isn’t to be had anywhere else inside the marketplace,” stated Angam.
Going Forward
At gift, Parallel Dot’s audience consists of organisations who need to leverage the power of AI to enhance their techniques. These organisations have lots of records. While some shape of analytics is in region because of the ongoing revolution in facts analytics for the beyond decade, many hidden insights can handiest be harnessed the usage of AI. “And as AI is turning into mainstream, we have determined firms to be extraordinarily receptive to this new generation,” he brought.
Also Read:- Lessons Learned Building a Startup Team
ParallelDot’s clients spreads throughout geographies and belongs to different sectors – consisting of finance, information and social media, organisation, etc. Due to diverse use-cases of their generation.
In the next 6-12 months, they want to continue building global-class generation as they discover to head deeper into AI. “We will retain innovating the prevailing merchandise and building new merchandise,” stated Angam.
Editor’s Note
According to Angam, the AI gift in the enterprise is still very nascent. It is most effective able to perform tasks that would take a human less than a 2d to complete. For instance, categorising an image, detecting objects in a photo, figuring out a sentiment in a text, and many others.
The principal advantage that AI has is that it is able to perform such duties across millions of documents in only mins. This is why ever-increasing data is propelling the want for smarter AI. And this need is only going to boom within the destiny and where predictive, intuitive solutions provided by using groups which includes ParallelDots are the want of the hour.
