🚀 The GradientJ Automation Platform is now Velos 🚀
🚀 The GradientJ Automation Platform is now Velos
🚀 The GradientJ Automation Platform is now Velos 🚀
How Humanly uses
Velos to create smarter, faster chatbots
How Humanly uses
Velos to create smarter, faster chatbots
How Humanly uses
Velos to create smarter, faster chatbots
With BJ Shannon, Chief Customer Officer at Humanly
50%
50%
50%
less time spent managing chatbot updates
less time spent managing chatbot updates
less time spent
managing chatbot updates
$100k
$100k
$100k
in new revenue generated through Q&A chatbots
in new revenue generated through Q&A chatbots
in new revenue generated through Q&A chatbots
2.5x
2.5x
2.5x
time saved on customer onboarding
time saved on customer onboarding
time saved on customer onboarding
Company
Humanly provides recruiting software that companies need to source, screen and communicate with talent
HQ
Seattle, WA
Industry
HR Tech
INTRO
Humanly (YC W20) is an HR tech company aimed at helping high-volume recruiting teams source, screen, and schedule candidates using AI.
Their flagship product is a chatbot that helps recruiters more efficiently and equitably screen & schedule candidates.
In this case study, we’ll go over how Humanly used Velos to launch a new product offering that companies like World Bank and Dish Network now leverage as part of their recruiting strategy.
Humanly (YC W20) is an HR tech company aimed at helping high-volume recruiting teams source, screen and schedule candidates using AI.
Their flagship product is a chatbot that helps recruiters more efficiently and equitably screen & schedule candidates.
In this case study, we’ll go over how Humanly used Velos to launch a new product offering that companies like World Bank and Dish Network now leverage as part of their recruiting strategy.
PROBLEM
Before Velos: Building customizable recruiting chatbots with AI proved difficult
Before Velos, Humanly used Azure’s AI services to power their AI chatbots using keyword matching. However, the process to set up keywords was not only extremely time consuming for their Customer Success team but it also wasn’t providing jobseekers with the best experience due to issues with accuracy and latency.
When OpenAI launched ChatGPT in 2023, Prem Kumar, co-founder and CEO of Humanly, knew he wanted to integrate LLM’s into their chatbot experience, and fast. The challenge was finding a partner with deep expertise in AI and machine learning who could operationalize this solution hassle-free without Humanly having to build out an army of machine learning engineers or data scientists.
“Whenever a new technology hits the market very quickly, there’s a lot of figuring out to do,” said Prem.
“Had we done it ourselves, it would have taken us plugging into multiple different LLMs. It would have taken a lot of work. Velos delivered the power of choice as a service, where we plug into their APIs, and can experiment with different models for different scenarios, while also being able to meet our customers' safety and functionality requirements.”
Knowing that this could be a key differentiator for Humanly, Prem went in search of a solution that would enable Humanly to use LLMs to power their chatbots and manage their customers' knowledge management process, as well as save time for both their Engineering & Customer Success team.
“With Velos, we can now give the power to our customers to allow their candidates to ask any HR/Recruiting question that might not be pre-programmed with an answer from the recruiting team, and instead can be pulled from their help documents.”
Before Velos, Humanly used Azure’s AI services to power their AI chatbots using keyword matching. However, the process to set up keywords was not only extremely time consuming for their Customer Success team but it also wasn’t providing jobseekers with the best experience due to issues with accuracy and latency.
When OpenAI launched ChatGPT in 2023, Prem Kumar, co-founder and CEO of Humanly, knew he wanted to integrate LLM’s into their chatbot experience, and fast. The challenge was finding a partner with deep expertise in AI and machine learning who could operationalize this solution hassle-free without Humanly having to build out an army of machine learning engineers or data scientists.
“Whenever a new technology hits the market very quickly, there’s a lot of figuring out to do,” said Prem.
“Had we done it ourselves, it would have taken us plugging into multiple different LLMs. It would have taken a lot of work. Velos delivered the power of choice as a service, where we plug into their APIs, and can experiment with different models for different scenarios, while also being able to meet our customers' safety and functionality requirements.”
Knowing that this could be a key differentiator for Humanly, Prem went in search of a solution that would enable Humanly to use LLMs to power their chatbots and manage their customers' knowledge management process, as well as save time for both their Engineering & Customer Success team.
“With Velos, we can now give the power to our customers to allow their candidates to ask any HR/Recruiting question that might not be pre-programmed with an answer from the recruiting team, and instead can be pulled from their help documents.”
SOLUTION
After Velos: Embracing a new era of AI-first candidate experience
With the help of Velos, Humanly was able to sell a new solution – an AI chatbot that provides prospects and applicants with streamlined conversations around commonly asked hiring questions that reduce mundane and repetitive questions and does it in a way where customer experience was at the forefront.
Humanly’s AI Q&A chatbot can draw from multiple sources of internal and external content and use it to create suggested answers to candidate queries, saving recruiting teams the time and effort of having to manually respond to hundreds of emails every week.
BJ Shannon, Head of Customer Success, oversees the team whose job it is to ensure customers can fully maximize their use of Humanly’s platform and receive a standout experience when seeking support.
“Velos has dramatically improved our ability to be agile and quick for our customers’ needs,” said BJ.
“Implementation is very simple, and the ability to include the same deployment across multiple chatbots (UAT/Sandbox, Production, etc) enables continuous testing for our customers even after we’re live.”
Looking ahead, Humanly plans to continue enhancing its AI Q&A chatbot operations with Velos by focusing on:
Ongoing optimization: Creating a monitoring system to ensure that the content uploaded by customers has the most accurate and relevant information and doesn’t conflict with other content in the customers knowledge base.
Expanding use cases: Exploring additional ways that LLMs can be utilized to support other areas of the business like resume parsing and scoring candidates based on chatbot questions and relevant skills.
With the help of Velos, Humanly was able to sell a new solution – an AI chatbot that provides prospects and applicants with streamlined conversations around commonly asked hiring questions that reduce mundane and repetitive questions and does it in a way where customer experience was at the forefront.
Humanly’s AI Q&A chatbot can draw from multiple sources of internal and external content and use it to create suggested answers to candidate queries, saving recruiting teams the time and effort of having to manually respond to hundreds of emails every week.
BJ Shannon, Head of Customer Success, oversees the team whose job it is to ensure customers can fully maximize their use of Humanly’s platform and receive a standout experience when seeking support.
“GradientJ has dramatically improved our ability to be agile and quick for our customers’ needs,” said BJ.
“Implementation is very simple, and the ability to include the same deployment across multiple chatbots (UAT/Sandbox, Production, etc) enables continuous testing for our customers even after we’re live.”
Looking ahead, Humanly plans to continue enhancing its AI Q&A chatbot operations with Velos by focusing on:
Ongoing optimization: Creating a monitoring system to ensure that the content uploaded by customers has the most accurate and relevant information and doesn’t conflict with other content in the customers knowledge base.
Expanding use cases: Exploring additional ways that LLMs can be utilized to support other areas of the business like resume parsing and scoring candidates based on chatbot questions and relevant skills.