Nanonets, a startup that uses AI to automate back-office processes, has raised $29 million in a new funding round led by Accel, aiming to improve the accuracy of automated processes involving large amounts of unstructured data.
Processing unstructured data from documents such as invoices, receipts, and purchase orders often requires repetitive tasks and significant human resources. Primarily targeting the financial services sector, Nanonets says his company's AI platform aims to improve the efficiency of these processes and make them more cost-effective.
The startup, an AY Combinator alumnus, has built an AI platform through which it offers no-code solutions that help companies extract information from documents, emails, tickets, databases, etc. and turn it into actionable insights. It is said that they are doing so. The company's AI platform uses machine learning architecture to analyze unstructured data and extract useful information from uploaded documents. Its no-code AI agents connect to ERP platforms like QuickBooks, Xero, Sage, and NetSuite to automate accounts payable processes, pull historical data from Square and Tableau to optimize supply chains, and extract data from patient management systems. You can summarize your health report.
Nanonets claims that manual invoice processing typically takes 15 minutes, but its automated finance solution can reduce that time to less than a minute. These solutions can be used for processes such as accounts payable, reconciliation, accounts receivable, and expense management.
The startup plans to use the new funding for research and development, improve the accuracy of its systems, and invest in sales and marketing. The company has approximately 100 employees, including most of its engineering team based in India. The company also plans to use the new funding to increase its workforce.
The all-stock Series B round was joined by Nanonets' existing investors Elevation Capital and Y Combinator. This brings the startup's total funding to $42 million, including a $10 million Series A round in 2022.
Prathamesh Juvatkar, co-founder and CTO of Nanonets, told TechCrunch that the startup will initially use convolutional neural networks, a neural network architecture used in computer vision for image classification and object recognition. He said he inspected the images and detected an object of interest. The startup then considered implementing Graph His Neural Networks, but ultimately moved to Transformers and adopted the multimodal Architecture, finding it more accurate than existing machine learning technologies. .
“Right now, we have multiple architectures on the backend. Every time we acquire a new customer, we train all these models based on customer data and see which one is more accurate,” he said in an interview. I mentioned it in
Juvatkar, an IIT Gandhinagar alumnus, and CEO Sarthak Jain co-founded Nanonets after selling Cubeit, a machine learning platform that turns web pages into shareable mobile cards, to fashion portal Myntra in 2016. Established.
Unlike many other AI startups that rely on large-scale language models (LLMs) or GPTs, Nanonets avoids the problem of hallucinations that occur when an AI system generates information that is not present in a particular document. I prefer Transformers. LLM knowledge.
Although the machine learning architecture used by Nanonets is document-agnostic, the startup is targeting the financial services sector since about 50% to 55% of its customers are in the financial services sector. We have provided a variety of integrations to streamline your financial operations. But the company is gradually expanding into “more adjacent processes” and has also started serving customers in healthcare and manufacturing, Juvatkar said.
Nanonets is not alone in the global market for AI-based workflow automation. The market is crowded with traditional optical character recognition (OCR) platforms as well as startups such as Rossum AI and Hyperscience. Large companies like UiPath also offer workflow automation, but they do so using structured data. Still, Juvatkar said Nanonets is up against the competition by offering his record 90% straight processing rate (percentage of data processed without manual intervention).
“We win deals primarily because of accuracy, user experience, and quality of integration,” he said.
Nanonets offers solutions in three different price points: Starter, Pro, and Enterprise. Of these tiers, Pro and Enterprise contribute the most to startups' annual recurring revenue, with an even split between them, Juvatkar told his TechCrunch. The startup also offers tools to convert PDF to Excel spreadsheets, CSV, JSON, XML and text, images to text, and images to Excel. The company says these converters have helped attract the attention of companies that need automation, reaching more than 34% of the world's Fortune 500 companies in the past two years. Additionally, the startup has grown its user base four times over the past 12 months and now has over 10,000 customers worldwide.
NanoNets has users all over the world, but the U.S. accounts for about 40% of its revenue, followed by Europe for 30% to 35%, Juvatkar said.
Juvatkar told TechCrunch, without disclosing numbers, that since the 2022 round, Nanonets' revenue has consistently increased 3x annually. The startup aims to double or triple its sales this year as well.
Consistent revenue growth is one of the reasons investors have invested in AI startups despite the global market slowdown. According to Tracxn, funding for AI startups has jumped from $10 billion in 2022 to $21 billion in 2023, but the number of deals has declined by 61 in the last year. He ended up with 399 cases. US AI startups received the most investment, followed by companies from China, the UK, Israel, and India.
“We are excited to partner with Nanonets on their mission to transform back-office operations with AI. Sarthak and his team are dedicated to fundamentally solving customer pain points and We've built a powerful solution that fully automates your processes end-to-end.What made Nanonets stand out to us is its comprehensive platform and Straight-Through Processing (STP) capabilities.These qualities make Nanonets stands out in the automation space and has already proven to have a positive impact on customers,” said Abhinav Chaturvedi, Partner at Accel. In a prepared statement.
Note: This article has been updated to reflect the Accel name change following a request for clarification from Accel.