KARTAVYA MATCH MAKER

Kartavya

Nagwekar

Abhi

Inder

31-05-2003

---

The Matchmaker

We want to make Global Recruiter, the ultimate matchmaker.

It should evolve to a stage, where it is able to tell,

▶ The Jobseeker

– Your chance/probability of getting this job is [___].

▶ The Employer

– Your chance/probability of selecting this candidate is [___].

Right now, we are quite far from that stage. But there is an opportunity of making a quick “beginning” in that direction.

The enclosed chart is almost self-explanatory.

What we need to do is to treat every job-advt. as a “document” for data mining — in much the same way we are treating every resume.

Then extract “function-related” keywords from 100,000 job-adverts, calculate their “frequency of occurrence”, and plot:

FUNCTIONAL PROFILE GRAPHS

– just as we do in Resumes.

We must use only those “keywords” which are found in 100,000 JOB-ADVTS.

Some of these keywords may be common with those found in resumes — but that is a different matter.

---

One question:

To begin with, how (and who) will assign a given FUNCTION to a given JOB-ADVT?

And this has to be done for 100,000 job-adverts!

Who would have time/energy to read job-description para in...

...100,000 job-adverts, in order to decide

“What IS the function of this job-advt?”

Can “Auto-Converter” tool, which Deepa uses, do this?

If that is not possible, then we may have to depend upon:

KEYWORDS FOUND IN RESUMES

(those keywords which define a given resume’s Functional Profile),

and then use these very same keywords to determine the FUNCTION of a job-advt.

This may not be such a bad idea, after all, considering that, in the ultimate analysis, we are trying to match:

Resume → with → Job-Advt.

Jobseeker → with → Employer

Where:

→ Resume is an “Agent” of Jobseeker

→ Advt is an “Agent” of Employer

And, we are trying to get these “Agents”

...to

→ “Talk” to each other (establish compatibility)

and then

→ “Negotiate” on behalf of their respective Masters/Owners.

To determine/establish:

MATCH-MAKING INDEX

The higher the Index, the better the Match — and vice versa.

And a recruitment manager can “preset” the INDEX, so that all resumes failing to reach that Index get summarily rejected/eliminated, in the first round of:

AUTO MATCHMAKING (SCREEN)

Then, the shortlist that emerges would call for human intervention/examination/matchmaking.

Click on the name of a Candidate

(in the shortlist developed thru AUTO MATCHMAKING),

would display both graphs side by side, as shown in enclosed sheet.

So, now the recruiter is able to see what is the “Degree” of match, visually.

“A picture is worth a thousand words.”

(Confucious)

A graph lends “credibility” where none exists!

There is something “mythical” about it!

And then by placing some appropriate “message” below the graph, we are trying to “influence” the (shortlisting) decision of the recruiter (in a subliminal way).

If he “bites”, we have a first-class

SELF FULFILLING PROPHESY!

If the candidate gets that job, both the candidate & the recruiter become “Great Believers” in the power of

THE MATCHMAKER!

They will swear by it & tell others too!

As far as candidates are concerned, the process will work in reverse.

First of all, his resume will have to be in our Online resume database.

Without this, he cannot conduct a jobsearch or receive a job-alert. (Today, this is not a condition on our website).

Because,

Only when we have his resume, can we plot his Function Profile Graphically.

After conducting a jobsearch, when he “shortlists” a few jobs, he would need to click on each job-advt.

This clicking will, once again, reveal the 2 GRAPHS, side-by-side

→ The functional graph of that job-advt

→ “ ” “ ” his resume

Once again, an appropriate message will appear below his graph, telling him:

→ You are an excellent / good / fair / poor “match” for this job.

→ You have “Good / Fair / Poor” probability of getting an interview-call against this job-advt.

Then, should he decide to click on

[ APPLY ONLINE ]

our server will forward his Butterfly resume to the advertiser.

But, if we have succeeded in developing such a scientific / probabilistic / personal analytics (–the more the jargon, the better!), tool,

(call it Business Intelligence – BI

  Artificial Intelligence – AI)

then, why should candidates even bother to “come” & conduct “jobsearches” on our website?

We tell them, “It is a waste of time.”

Just enter your Search Parameters in

“REQUEST FOR JOB ALERT” screen

(Ind / Function / Design Level / Posting City, etc.)

And,

  also select

MATCH-MAKING INDEX

Then sit back & relax.

We will deliver matching jobs to your doorstep!

Such “Match-making Graphical Displays” have the power to pull in MILLIONS & MILLIONS of resumes to our website!

This would happen, even if we do NOT offer to send out “job-alerts” to candidates.

They would find time to log into our website & endlessly conduct job-searches among 500/1000 job-adverts that Deepa will download everyday (using Myriad) & upload on our website.

At any given point of time, there will be 20,000 / 25,000 job-adverts on our site.

[Of course, by end Aug., we will need her to first download 100,000 job-adverts in order to develop keyword-based function profiles in the first place.]

Nothing can start until then.

“Candidate Alert” feature (to corporates) should wait till we have accumulated 500,000 resumes.

---

I strongly feel, Match-maker could be our “lucky-break.”

It is only a matter of getting “ready” – a question of our “preparedness” to seize that opportunity – before someone else does.

Let us go out & do it. I am confident we can & therefore, we must.