Can AI help you find the right talent according to the specific needs of your organisation? A startup in India thinks it can.
Enterprises often face challenges such as talent shortages in specialised fields, fierce competition driving up salaries, and the need to ensure new hires fit the company culture.
Additionally, evolving skill requirements– especially with the advent of generative AI– complicate finding candidates with the most current expertise.
Hence, in the fast-evolving world of business, managing your talent pool efficiently becomes key. Spire.ai, which is based in Koramangala, Bengaluru, has developed an AI-powered assistant that helps large enterprises better manage and develop talent.
The AI assistant automatically identifies skill gaps in an organisation, performs skill mapping, provides career recommendations to employees by scanning granular data, and predicts future skill gaps.
A Copilot for Talent
“Consider a professional services company with over 100,000 employees. These companies frequently rotate staff across various projects for their, let’s say, 5,000 clients. They face a constant challenge in aligning supply with the high demand from clients,” Saurabh Jain, founder & CEO at Spire.ai, told AIM.
To manage this, they address the supply-demand balance in several ways– building versus buying talent, rescaling excess supply, commoditising or monetising surplus, and internal resource management through bartering or lending.
“For instance, suppose an organisation needs engineers skilled in React.js, but you have experts in Angular.js; those AngularJS professionals might still be valuable. They could potentially be trained in React.js within one to two weeks,” Jain said.
The AI assistant leverages the enterprise data and automatically provides this suggestion to the enterprise and also lists out the engineers who excel in Angular.js and can be reskilled for React.js.
This saves the enterprise a lot of time and effort. Instead of hiring new talent for one project, they managed to meet the needs internally. This has a direct impact on the enterprise’s operational margin.
“This has to happen in real-time and on a continuous basis within organisations. Imagine if you’re doing that for 300,000 employees globally across, let’s say, 60 countries. You have to continuously know all of this. So for this you need intelligent systems to do that,” Jain pointed out.
Currently, the available systems are workflow-based, requiring enterprises to input workflows for the system to operate effectively, which, according to Jain, is a single point of failure.
The AI assistant also analyses unstructured data from various sources, to offer skill inventory and profiling solutions and marketplace and talent cross-pollination.
The Secret Sauce
At the heart of Spire.ai is a Large Graph Model (LGM) featuring over 10 million sanitised graph nodes. This model encompasses skills across 26 industries, including tech, healthcare, retail, supply chain, defence, and pharmaceuticals.
The LGMs process extensive data, such as job descriptions, industry trends, and learning materials, to automatically identify and generate various roles and the complex skill sets needed for them across any industry or business function. This self-evolving skill framework removes the need for manual mapping.
“On top of the LGM, we have developed the Spiral Bot Engine, which powers our applications and focuses on search and match capabilities, integrating with data management, data inference, and interpretation layers. It supports data interpretation in over 250 global languages,” Jain revealed.
However, interestingly, the startup has not integrated Large Language Models (LLMs) into its solutions. The LGM, coupled with the Spiral Bot and the application and insight layer– which is also part of the architecture– is substituting an LLM right now.
But they are not ruling it out, either. “There is a good amount of interest from the industry that, unlike a regular LLM chatbot, where the user invokes the engine, can the engine invoke the user and then carry the conversation? That is the last leg that we are working on,” he said.
Customers
The startup’s product is designed to solve hiring challenges for large organisations but can also be used by smaller and midsize enterprises. According to Jain, any organisation with over 500 employees will find Spire.ai’s AI assistant useful.
Moreover, for smaller enterprises, the startup plans to launch the solution in an ‘application mode’ this year, specifically designed for smaller enterprises whose employee count is less than 500. Currently, it serves large IT enterprises, telecom and communication companies as well as insurance companies in India.
