Matching Technology

talent peak matching technology can be used to make recommendations which can help program leads form a team of matching mentors and mentees since the matching criteria can vary from program to program, talent peak matching technology provides a lot of flexibility to configure the matching algorithm that matches the requirements of each program setting up the matching algorithm the team matching algorithm can be configured from choosing ‘teams configuration’ and choosing the ‘matching algorithm sub tab the algorithm allows use of any existing employee attribute that can be provided by hris system (e g workday, sap, adp) or custom attributes that can be collected from the user as part of the program join request the above screen shot shows an example program configured with primary attributes that were received from hris and custom attributes that the user will provide as part of their program join request survey each of the attributes can be configured to match or not match or be ignored the custom attributes can be configured by clicking on “ + ” sign next to the select custom attributes you want to match section which will open a survey creation form as shown below the custom attribute survey form allows data to be requested in the form of multiple choice question which is depicted with a check box (see fav color in the example below), single choice question which is depicted with a radio button (see fav sport in the example below) or rating (see negotiation skill level below) application of the matching algorithm once configured the matching algorithm is put to use on the talent peak application either when the program lead is trying to pair mentors and mentees for admin led programs or when a user of the platform is trying to discover matches in peer to peer programs peer to peer mentoring the matching algorithm compares the attributes of the subject mentee with a pool of available mentors to make recommendations of the best match similarly the mentors can also be matched against the available mentees the matching recommendation is provided as a numerical score which is calculated as the average of %age matched of each attribute the percentage match for each attribute depends upon the attribute type for single choice questions or numerical/rating attributes, percentage match is either 0% (not a match) or 100% (a match) e g for fav sport above, if mentor and mentee both select soccer then it is a 100% match but if mentor selects soccer and mentee selects basketball then it is a 0% match for multiple choice questions, the percentage match is calculated as number of matched options divided by the number of options that would give a 100% match for example if mentee selects red and yellow as their favorite color and mentor selects red as favorite color then it is considered a 50% match as one out of the two selected options matched for every recommendation provided by the matching algorithm, the details can be seen by clicking on the “i” button next to the overall match (as shown below) participants will be able to see matches through the discover tab on their my teams page this will present the user with an array of possible matches for their peer to peer program a user can then use the send request button to request partnership with that mentor/mentee requests will be sent to the requestee's email as well as live on their requests received tab users can see and manage their requests from the requests sent tab please note, the matching algorithm is just a recommendation tool and it is up to the discretion of the program lead to use the recommendation or not sometimes program leads download the survey report to look at the raw data and form teams based on their own criteria ai team builder admin led pairing once registrations have been compiled, program leads can use the ai team builder to create teams using the matching algorithm configuring the ai team builder under the manage teams main tab, select the manage teams sub tab and locate the ai team builder upon clicking on the ai team builder button, you will be presented with the ai team builder configuration screen team naming convention this first step allows the program lead to indicate how the teams will be named there are two choices member names this will name each team using the names of the members such as john doe & jane smith program leads can choose the connector between the two names the options are dash > john doe jane smith ampersand > john doe & jane smith slash > john doe / jane smith and > john doe and jane smith prefix + sequence number this will name each pairing with a prefix (team, partnership, cohort, etc ) and a number, starting from the number indicted sample team names are team 23 partnership 145 cohort 9 team composition the second step is to create the team composition; how many mentors and mentees should there be on a team? simply indicate the number of mentors and mentees you would like the team to be comprised of using the appropriate box match quality this third step indicates the lowest possible percentage match threshold a team has to make to be created this can be set from 0% 100%, but a higher match percentage may result in fewer matches made by the team builder click the start matching button to run the team builder making matches once start matching is clicked, you will see the team builder start to make matches if all mentors and mentees are matched, you will see a listing of the teams created if there are mentors and mentees who do not meet the match threshold, you will receive this notification which also indicates the number of mentors and mentees left the remaining mentors and mentees can be easily matched using the manual matching panel simply drag and drop the remaining mentors/mentees to create teams click the +create match button to add the match to the list of matches reviewing matches once matches are made, they can be reviewed using the matches panel click the checkmark to accept the match click the red x to deny the match click accept all pending to accept all matches click revert to pending to change all matches back to pending click reset all to start fresh when all matches have been made, click accept all pending at the top of the tool, you will have some additional choices click resume to go back to the builder click download to get a speadsheet of matches made click create all teams to allow all matches to be made click exit to leave the team builder activating teams once create all teams in the team builder has been clicked, the teams will be created in the manage teams tab make sure you're changing the teams status filter at the top of the match listing to ensure you're looking at draft teams teams can individually be managed by clicking on the associated action button edit allows the team to be modified manually activate makes the team active which will send out activation emails and count as day 1 for the match delete this will remove the match teams can also be managed in bulk by selecting select bulk action bulk actions will perform the change indicated on all teams in that teams status choosing "change all draft to active" will send out activation emails and count as day 1 for the match