With the advent of AI language models and their rapid evolution, researchers and academicians today have access to several tools to ease their work. Despite this, literature reviews remain a challenging task, eating into much of their time and energy.

To resolve this, two Indian founders Saikiran Chandha and Shanu Kumar are determined to build solutions, using the power of generative AI, to automate the task of decoding and understanding research papers for the scientific research community.

Chandha, who is a product of Vellore Institute of Technology (VIT), and Shanu Kumar, an IIT Kharagpur alumni, first met at a hackathon.

“At the Stanford Ignite program, we were trying to figure out the problems that we face daily and find a solution that could be viable from a business point of view. As researchers, we both clearly understood the pain points that the scientific community and academicians go through in their daily work. The resolution of this problem was up our alley,” said Chandha, the cofounder and CEO of SciSpace, a Bengaluru-based software development startup.

According to the company’s website, SciSpace was founded in 2016 and specialises in creating SaaS platforms for academic and scientific research. However, according to the CEO, the startup went through a major transition with the advent of generative AI, and the launch of its flagship product, Copilot, three months ago. SciSpace’s Copilot is an AI research assistant that empowers researchers to decode any research paper effortlessly.

An AI-Powered Research Assistant At Your Service!

Speaking with Inc42, Chandha, said that while there are several platforms, including Google Scholar, which help researchers with multiple links to research papers, a lot of human effort and time is still required to figure out if they are relevant to one’s research.

With its tool, Copilot, SciSpace not only provides academicians and researchers with relevant links to research papers but also summarises them to save time and effort. Moreover, users can add papers of their own choice to extract data. In the chat box, users can ask the tool to specify the content of the documents, and explain certain paragraphs, the whole paper, or specific portions.

SciSpace breaks down any complex problem into sub-tasks and sub-problems and helps build models for each sub-tasks or sub-modules by using open AI and Anthropic as foundational models. To top it up, the startup has its proprietary models that help to fine-tune these results.

These proprietary models have also facilitated SciSpace in building capabilities beyond just providing summaries. For context, if users are unable to understand a math problem or an equation, the SciSpace Copilot comes in handy.

Right now, the platform has 250 Mn open-access research papers on its platforms. Moreover, it launched a Chrome plug-in a month ago. Going forward, the generative AI startup wants to expand its capabilities beyond research papers.

“We will start indexing white papers, case studies, podcasts, and videos. We want to cater to users beyond the research community,” Chandha said.

Banking On The Perks Of A Freemium Model

The startup, which claims to have 4,00,000 active researchers on its platform and 1,00,000 users on its recently launched Chrome plug-in, does not charge a single penny from its users, as it wants to solidify its data repository. 

“The decision is helping us to collect data and fine-tune our model,” Chandha said. However, plans are afoot to make way for a revenue stream for a sustainable business. Further, the startup is also looking to charge organisations involved in a significant amount of R&D work. 

Currently, the startup is involved in doing pilots with two life sciences companies in the US. It also plans to utilise its expanded capabilities for monetisation.

SciSpace has raised an undisclosed funding amount from VC firms such as Inventus Capital and Silicon Valley Quad.

Moving on, the startup counts academic search engines such as Google Scholar and ResearchGate and citation managers Zoreto and Litmap as its competitors, which cater to the same audience. However, none of these platforms offers data extraction from research papers right now.

Amid the generative AI boom, SciSpace stands at a sweet spot as it reduces the time and effort of its users but its business viability still very much depends on how users will respond to monetisation plans. 

Moreover, big tech players are evolving quickly to respond to the emerging demand of users, reaping the maximum benefits of generative AI. It is this breakneck evolution and adoption of AI that could increase competitive headwinds for smaller startups such as SciSpace going ahead.