17
Feb 2004
OES,
Module1, Non Member Database.
To,
Nirmit/Raju/Sriram/Mitchelle/Sanjay/Amol/Reena
This
is regarding OES, Module1, Non Member Database.
I
have following suggestions to make. I request all of you to go through this
document, debate over it and reach to a conclusion. Our decision will decide
what kind of tools/software will get developed in future in 3P.
MUDULE 1
Suggestion
1: Let us away with the concept of Industry.
Why
:
Since
in past we found that DEOs (Data Entry Operators) are not really capable of
defining the industry, we developed a tool using which, depending on the
company name we built a knowledge-base of Industry. While adding a new resume.
If the current company is present it the knowledge-base then industry gets
defined automatically, otherwise it remains BLANK. Because of this that particular resume
becomes un-searchable on Industry parameter. One has to check all such resumes
on periodic basis and keep on refining/updating the Industry Knowledge0base.,
which is difficult. So this logic is not full-proof.
Suggestion
2: Let us get away with the CURRENT concept of Function and Designation
Level
Why:
Here also the problem is same. We have
developed a tool, which stores the Actual Designation and splits it into
Function and Designation Level. While adding a new resume, if the current
Designation is present in the knowledge –base then Function and Designation
Level gets automatically defined, otherwise it remains BLANK. Because of this
that particular resume will not be searchable on Function as well as Designation
Level Parameter. One has to check all such resumes on periodic basis and keep
on refining the Knowledge-base, which is difficult. So this logic is not full
proof.
Here
I have assumed that the Function is the first parameter to decide if a
candidate is a right fit and then comes Industry. Designation Levels are
ambiguous in practical world and change from company to company.
Solution : LET US GURUMINE & GURUSEARCH
Why
:
In
last 12-18 months we have done a reasonably good job in developing GURUMINE and
GURUSEARCH. Till date we have added around 1.25 lakh resumes belonging to 3p in
the database. Remaining can be added within 40-60 days, at max.
Let
us for argument sake; assume that the extraction is not ACCURATE. What we can do
is, on daily DEOs will extract the resumes in bulk and next day they will
verify them. By this method we can eliminate the problem of accuracy. Similarly
whenever they want to add a single resume they can add it quickly.
We
all know and believe that the Function Exposure part of GuruMine is almost
accurate, no need to be dependent on anyone to assign the Function. On top of
it GuruMine assigns 3 functions not only 1.
While
storing the full resume, one can directly store them in EXECITIVE BIO-DATA
Format, because of which consultants will be less dependent on DEOs while
sending the resumes to client.
We
can even avoid the duplicate resumes while extracting, which is the biggest
advantage.
Today
Member and Non Member are two different tools. Where as one can add a record to
GuruMine even if we have no resume of that person. Next time when we will get
the resume of that person we can update the record.
As
far as backup is concerned, since we are storing the full resume to the
database itself, only database backup is more that enough, which takes less
time and space compared to current backup.
Since
Module 1 and Non member tools are based on client-server technology, they are
really difficult to maintain. Whereas GuruMine and GuruSearch use the
Internet/Internet based technology, the maintenance is centralized like OES.
Additional
Work to be done to use GuruMine & Gurusearch
If
we decide to use GuruMine and GuruSearch, following things have to be added / modified
1.
We
have to build interface between GuruSearch and OES. (Shortlisting of resumes
for a given assignment and passing the data to OES)
2.
We
have to add Actual Designation based search to GuruSearch which will be similar
to Keyword Search.
4.
SQL
Server 2000 has to be installed and maintained, which is much easier compared
to Oracle.
Negative
Side: All the PEN Nos will get re-assigned / changed.
I
have tried sincerely to put-in my serious efforts / thoughts to analyse the
situation. I am sure you people can do a better job than me and suggest a
better solution / way to tackle this problem. So let us do it once and for all.
Hemen
Parekh
No comments:
Post a Comment