OpenInterview
Problem: There is a dwindling pool of talent and a ballooning pool of applicants in every hi-tech company. There is an infinite cost penalty for recruiting the wrong candidate and near infinite value unlocked for a great hire!
Bottle-necking causes delays and time-to-market(hence revenue) loss.
All these factors
call for automation of the recruitment and associated processes.
OpenInterview solves several of these problems with AI, NLP & Cloud.
Solution: VideoBot for conducting the interview itself. We have user customizable interview content (initialized with public material from Google) for a few interview types. We use Cloud based Speech-to-text and Text-to-speech micro-services to give a seamless Human Computer Interface.
ML/DL based NLP techniques to analyze the answers. We used to use Cosine, which we upgraded to Lucene and now we are experimenting with BERT for semantic matching and scoring.
We also have a coding skill test and automated code scoring mechanism.
Bottle-necking:
Interviews can be scaled infinitely with Cloud and Bot Technologies making it possible to hire en-masse if required. So, we would have crossed the human 8 hour/day bottleneck which would typically mean 8 interviews/day/panelist max.
This can be achieved using AI.
Kaizen:
As deep learning and other models improve over time with incremental data and training, we expect accuracy to improve tending towards the most desired candidate. Hence we will implement a constantly improving "Auto-Kaizen" system in effect.
Bias:
By eliminating Human Factors in the selection process, especially for jobs that do not need or value the human element, we plan to achieve building great organizations for our customers. We also encourage our customers to use this tool for internal jobs, as well as a qualifying criteria for promotions.
Hiring Costs:
Huge direct costs of hiring caused by time spent by hiring personnel, both tech and non-tech, can be saved. Indirect costs incurred due to factors such an untrained hiring staff causing hiring mistakes. Costs incurred due to hiring mistakes and potential time lost in product development etc.
Customers:
Target customers initially are hi-tech companies. However the framework has the potential to support other industries.
Recruiting is
about half of every great technology company's problem! The value is near
infinite and we plan to make this a competitive differentiator for
our adopting customers.
Target Customers are
Hi-Technology Companies who are end recruiters of talent in India, Israel,USA and other favorable geographies.
I have 2 paper publications in this space. Prioritized Requirements were customer validated, we have involved the customer in product development (requirements definition and feature prioritization) MVP was ready in May 2019. We are now getting ready for Beta with 2 customers early 2020.
We endeavor to provide a:
"Video interview bot and code evaluation SaaS platform that automates talent acquisition for the hi-technology industry”
OpenInterview's vision is to "Code till singularity!"
I believe later versions of the platform will yield highly repeatable and accurate, fully automated, talent screening and selection tools and technology.
As for the population of interviewees in question, the tool will ensure standardization in tests, consistency in difficulty levels and spread of material, an automated process that guarantees fairness to a large extent. The platform will also provide a seamless Human Computer Interface to the interviewees and a consistent and bias free communication experience during all parts of the candidate selection process. Bias and prejudice will not be designed in the platform and the platform will have a "no tampering" policy with the universal past data, which will become a single version of truth on which future decisions will be based, whatever that may be.
- Upskill, reskill, or retrain workers in the industries most affected by technological transformations
- Support underserved people in fostering entrepreneurship and creating new technologies, businesses, and jobs
- Prototype
AI and NLP with BERT and more advanced models in the future with Reinforcement Learning components.
Cloud technology for near infinite scaling and green computing
Automated evaluation, taking out human issues.
Not sure
- Women & Girls
- Pregnant Women
- Children & Adolescents
- Elderly
- Rural Residents
- Urban Residents
- Very Poor
- Low-Income
- Middle-Income
- Minorities/Previously Excluded Populations
- Refugees/Internally Displaced Persons
- Persons with Disabilities
- Israel
- United States
- Israel
- United States
Year 1: Two signed up Global Beta Customers
Year 2: Upto 7 Global players
Year 3: All key hi-tech players in the World
We plan to build meritocracies for our adopting customers.
This automated tool will also create fully standard compliant processes for talent selection thus helping meritorious candidates get picked to jobs they wish to apply to. Thus this tool will end up creating a bias-free world for all.
Barrier 1: Technical Talent in the Cybersecurity, Web/mobile development and other spaces to constantly secure the product. This cannot be overcome quickly with capital Infusion, though it can be eased
Barrier 2: Patents for Automated Interview technology, which I was working on were first applied for and obtained by an extremely large competitor. However, we believe we do have freedom to operate based on my published research.
Barrier 3: Legal and compliance issues wil change per geography, creating fragmentation in the unified cloud SaaS platform
Barrier 4: Cultural factors such as selection preferences may vary by Country/Region. Again this will prevent evolution of a single unified Machine Learning Model for making choices
I am thinking about these issues at this time
- My solution is already being implemented in one or more of ServiceNow’s primary markets
- I am planning to expand my solution to one or more of ServiceNow’s primary markets
Please supply NDA and Non-Compete Agreement to access this information
Please supply NDA and Non-Compete Agreement to access this information
- For-Profit
n/a
3-4 full time
2 part time
Domain Folks on an on-demand basis
I am not sure about this. I have not surveyed the rest of the world
For-profit global tech players
this material is already covered before
Initial 2 customers will be served in bootstrapped mode. Capital raising rounds will follow later
Got the link through MIT technology Review, and decided to apply. I think I have a shot at winning 100k but, the least I will get is feedback
- Technology
- Legal
- Media & speaking opportunities
n/a
n/a