Is your Installed Base data a Mess?
Clean and unify Installed Base data with Entytle
You Might Also Like
Let me go ahead and introduce our speaker of the day.
We have Shri Moondata, who has worked with in titles
for the past eight years.
He takes care of the entire data and DevOps infrastructure
for in title.
He has been instrumental in building
an in-salt-based data studio focused on industrial data
workflows, which he's going to go ahead and jump into
with that being said, Shri, all.
But you take it away.
Thanks a lot for the introduction, Colin.
Hello, everyone.
And thank you for joining our webinar today.
I'm excited to talk to you about a common challenge
based by industrial OEMs, which is managing the install
based data effectively.
Most of you would agree to this that this is a comprehensive
process.
And with the next few minutes, I'd like to shed light on,
how do we do it differently and repeatedly, and at a larger
scale at the end of the title?
So let's start with the scenario, which many of you
might be familiar with.
Your install based data, as you see in this slide,
would be scattered across various functional systems,
like CRM systems, ERP, FSM, different spreadsheets,
or e-commerce tools, legacy systems, and so on.
And each of the systems may hold valuable information,
but accessing and unifying this data can be a daunting task.
So it's like having all the pieces of the puzzle,
but not being able to see the complete picture.
And this is what we see across every single prospect
or customer that we talk to.
On top of that, the data here would be often messy,
inconsistent formats, duplicate entries, incomplete records.
It really makes it difficult to gain accurate insights
on to the data.
It becomes a big hindrance and leads to missed revenue
opportunities.
And this problem gets further compounded by the data silos.
And it's not that companies do not
make investments in this area, but thing
is the lack of a comprehensive process
around this data-fitting cleaning notification
leads to eventually get another silo.
So what's the solution?
And where does the entire comment to picture?
So what the entire list has done is
we have created a clean and unified process
that does the data cleaning enrichment analysis
on your IB data, which keeps your IB objects at the center.
So if we look at this data journey,
so here what we are striving is to transform your raw data
into actionable insights through the entire installation
platform.
So we begin by gathering the relevant data
from different systems.
Then we move to data exploration, where
we do a lot of everything on the data quality
and the data cleaning part of it, before feeding it
to analytics pipeline, where we have built certain custom
purpose-built models centered around the IB data.
And only when we have validated those results,
practice, set them, we push it to the final destination,
which are basically actionable intelligence, which
are accessed by our customers leading
into more new opportunities for them.
Having said that, the most crucial aspect in this journey
is the data quality, as it lays the foundation
for accurate analysis and generate
reliable insights in your IB data journey.
So we'll just deep dive into the data quality piece
because it has to be really comprehensive.
So if you take a step back and understand
what are the primary objects around the IB data?
So they could be transactions, equipments,
item master, bill of materials, service data, service
contact data, warrant these location masters,
and all of these data in real life
is scattered across different systems.
So the first thing is it should be cleaned and unified
and also enriched further so that you can
trick this into a high quality install base.
And during this process, it's also
important to have proper data governance and security
practices because it's important to ensure
integrity and protection of data.
But having said that, it's a very recent and done,
basically.
As you can see in this slide, cleaning and stitching
this install base data is a comprehensive process,
but it requires a lot of investment,
diverse skillsets, multiple tools, which
is really a time consuming and expensive thing to do.
So at the entitled, we have built something called
as a part of the installation platform.
We have built something called as the entire install base
data studio, which unifies all the data from all of your
different systems into a single cohesive system.
So this ensures that we achieve high quality data,
focused around your IB use cases, and on that high quality
data, we generate C60 degree analysis
like the parts of your training, equipment
training services, revenue training, identify the health
of your customers, both at a macro and micro level,
and make specific predictions and prescriptions
around equipments, opportunities, parts, opportunities,
and so on.
So at a high level, this is what our approach is.
What I do is I directly jump into our product demo
because the proof of the pudding is in the eating,
where I would actually show, what is it that entire install
base data studio has to offer and how do we do it?
As you can see here, into the IB data studio,
we have different workspaces for different personnel.
One is data prep workspace, where we do all the data
profiling, stitching, and cleaning, and stuff.
There is the analytic and the setup, basically,
the meta data management.
In the analytics, we have descriptive, predictive,
and predictive opportunities on the clean and unified IB data
that would have done into the data prep workspace.
Once we jump into the data prep workspace,
you would see a lot of services.
All of them are like self explanatory.
So data prep is where you prep all the data,
data loader, and data load is where we ingest this data
into our target systems.
And deduplication is where we deduplicate locations,
which is a common identifier across all different systems
that we have spoken about earlier.
So on the data acquisition side, you can upload file
from your local machine, or you can import directly
from your databases, import directly from snowflake.
And once you have imported data, we run several jobs
to profile that data.
So if I jump into one of the jobs, what you can see here
is a data distribution, which helps in identifying outliers,
missing values, the patterns in your data.
Or example, if I look at the report of this particular data
center that I have, it has roughly 150,000 records.
And 30% of the data is missing.
Now, if you look at this, you have different attributes,
cardinality, or different columns, which denotes
what are the, what is the, what how many number of records
in your data are distinct?
It will buy the book, why was it your book,
so you would identify any outliers in your data.
So doing this in the initial phase helps you really identify,
helps you identify how does my data quality look like?
Is this because if take a classic case where your data
may be lying into two to three different systems,
we could unify that and bring all of that data together.
And this gives you a unified picture of your data
distribution.
And then you can take a call on how do you improve the quality
of this data?
And if it is good, you move ahead, or if not,
you can perform several elementary and complex data
operations.
Some of them are merging a column,
was splitting a column, deleting it, renaming it
for better readability.
There are no code-local operations like find and replace,
find and replace by regular expressions.
You can just find by certain pattern and then replace it
by a specific value.
There are SQL type joins, where you can do left, right,
inner join on your data.
Then so I can choose a data set and do a left, right, inner join
also out a join with any other data set.
There are no code operations like create new column of type
string number, date, Boolean.
And you can perform various operations.
So if column A is empty and column B is not empty.
So different kind of no code things.
You can define a uniqueness criteria on your data.
You can say that a combination of n number of columns defines
my record as unique.
And what it is, it would drop duplicates.
That's how you keep unique records.
Then there are other operations like SQL type group
by or SQL type ranking functions, where you can group
by certain things and then you can use certain functions.
One of the use cases is you would have
caught some initial set of data from your system.
Now you want to apply an additional data to this.
So you can union it pattern matching
barricular expressions.
So you can count and extract matches
by certain regular expressions.
And the best part is whatever you do,
you can put this as a part of the recipe so that the next time
when you get similar data, you can rerun the same recipe step
that you had written earlier.
And what you are seeing here is something
called a ribic continuum, basically, which
is a connected flow of different objects in a data set.
So I've got all of these transaction data
from different sources.
I have done a lot of cleaning that is common.
I've always objects in a single recipe
and then split into its respective destinations.
And these are the standard objects, like orders, assets,
services, service contracts, warranties, which primarily
deal with your IB use cases.
Now that this data would have been clean and enriched,
we also have a de-duplication service
where we can de-duplicate certain those locations
by whatever definition.
So you may say a combination of customer names,
the exact combination of location unique,
or just a combination of customer name
makes my accounts unique.
So as you can see here, this connects communication,
it is present across different systems.
And then we have brought all of them
as a part of the single cluster.
So this is how you de-duplicate your purchasing entities
or locations which help you unify your data
in a much better fashion.
Now that this cleaning and unification
will be done, we feed it to our analytics pipeline.
So we have a lot of services which work on this clean and unify
data and for better readability.
So I'll just show the configuration screen
where we have services both at a macro and a micro level.
At a macro level, we can generate what
is the propensity or probability that a location or transit
either part or an equipment or a service from you
in the next and number of days.
So once you train a model, you can use it for next and number
of days or you can train a new model based on data systems.
If based on the transaction history that you have got,
you can rank contacts by assigning a page
to their occurrences into different sources of the data.
And you can rank contact this, it typically
solves your whom to call problem.
Once you have opportunities created on the IV data.
We also have predictive and prescriptive opportunities
on the transaction history.
So you can generate opportunities at a micro level.
We would say that location X would transact item Y
in the next 90 days and this is the quantity and this is the
probability.
So this is and we call it as a purchase rate service.
At a micro level, we can segment your customers by usage
and the silver tellumer of parts and total number of
revenue that they would have generated.
We can also calculate the frequency, consistency and
recency of every purchasing entity or location.
What is the frequency between which they are buying?
What is the consistency?
And based on these metrics, we can also calculate
are your customers drifting away from you?
While a customer that was consistently buying from you has
that customer drifted away from you recently.
So these are the kind of information analytics that we
can generate at a macro and a micro level on that clean and
unified data.
So this is the USP of entire digital basically the studio
where we do all of this heavy lifting.
All of them under a single roof as you can see in this slide.
So you get data from different sources.
You perform, you do all the data stitching cleaning
unification.
We perform different purpose with analytics like calculating
a wallet share, a trend, propensity, and the predictive
and prescription opportunities that I spoke about a while ago.
And all of this can be viewed into a flagship product
which is the entire inside which gives you a 360 degree
visibility on your installments data where you can do
search or filtering or prioritize your hunting list.
You can also push this data directly into your CRM or into
data lakes.
We have customers who have given us access to their CRM
systems where we directly push it to their sales force or any
other system that would have configured.
We also have a lot of open APIs wherein our customers access
this clean and unified data via APIs.
And gain the value or basically derive maximum value on the
skin and unified data.
And one such example is how effortlessly we can integrate
this data into a sales force CRM.
So what you see, what you heard from me in the previous slides
before the past few minutes is all your major IB objects,
like equipments, parts, services, service contacts.
Once you have clean, enriched and unified that into the
entire install based intelligence platform, all of them can
be pulled seamlessly into your sales force account.
Thereby giving you a 360 degree view under a single roof
and giving you more than any opportunities and you can do
and there are endless possibilities to do things around
your install based data.
So this is what I had.
But today in the next webinar, we would be talking about
integration with sales force and also how
entitled analytics help our customers generate more revenue
by means of more predictive and prescriptive opportunities.
So with this, I would like to leave it to the audience for
any questions.
Yeah, we had a couple of questions come in.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
So I'm going to ask you a question.
And then we have another question here.
How easy is it to integrate and title with sales force?
And what does the add value it can give?
Yeah, so.
So it's a again a very good question.
And that's as you can see on my slide.
This is what you see once you have clean and identify data into your sales for
the account.
Where under that.
So under this I be intelligent staff just created by entitled.
You would see a 360 degree view of all your IB objects under a single tap.
So it's once the data is cleaned and unified.
The next step is pretty easy.
And we are basically we are very easy for our customers to have all of the data
under a single.
Right, they're keeping up with their current work flows that they have or the
current text set that they use.
But have the capability of expedited account research done all in front of them
Okay, that's all the questions that we had to come through today.
If anybody has any sort of follow up questions, please feel free to reach out
to us.
And we can get you connected to the right party here at entitled.
Right, with that being said, thanks everybody.
We look forward to seeing you in the next event.
Thank you.