Hello everybody, I am Lin.
I am an agri business development exec and have learnt basic Knime. I know basic nodes but that doesn’t really add up to anything as yet. I am not tasked with implementing data sci for my small company and I am quite lost suddenly. I would be glad if someone could point me to some links for a small company to track raw material pricing or manufactured goods. I have been looking up and down all day. Grateful 
Hi @Needpractice (Lin), and welcome to the KNIME community. I’m guessing you meant “now tasked” rather than “not tasked” 
I couldn’t tell you much about the raw material price tracking etc but I wonder whether having a look at some of the case studies here would give some useful insights or ideas:
There is also an upcoming webinar that may be of interest
I’m guessing you are in the situation where you have just discovered this amazing “multi-tool” and are wondering how you can use it to build something. The trouble is… what to build?
I’d probably start by looking at what kinds of data your company currently holds, and do a bit of brainstorming of ideas on what things you think you could do with knowing so that you can start transforming “raw data” into “valuable information” and “insights” to help move the company forward.
Start small and maybe ignore KNIME initially as you think about how you would achieve some simple goals manually. Once you have something to aim for, look to see if and how KNIME can help.
If you haven’t done so already, I’d suggest looking at some of the self-paced training courses available as these can help expand on your knowledge of what KNIME can do.
https://www.knime.com/knime-self-paced-courses?pk_vid=0f6ced8097f9bbec169849851684dce5
5 Likes
I would completely ignore KNIME and instead approach these questions from the strategic business side. Define you goals.
What data driven management decisions need to be prompted, what areas require better controls and tracking, what data integrations are required for a full picture or simpler processes, what areas impact profitability most, which changes would be easiest and most impactful initially, who would the final user of final reports or data be and how would they use them?
6 Likes
Dear takbb,
Thank you very much for your reply. Ok I will look at the training courses. I am already very worried about not getting a node to work. Knime is not idiot proof I must say. Like if I have something like that:
Date Price
29 Oct 23 USD 1
30 Oct 23 USD 1.1
31 Oct 23 USD 1.2
I can make a line chart with the above data ( I am already hearing office staff saying “but you can do that with Excel?”, to which I will tell them that Knime allows you to just import a dataset, execute and line chart appears )
I can also have append another column to the data above using Math Formula node, to calculate the price difference for each day. Except that what would be the formula to key in
(I tried and tried but failed, this is where Knime is hard for beginners, my python is not great and chatgpt wasn’t useful probably my prompt was weak).
Yes I agree. Our business is a very standard business model, it tracks production, quality, sales, delivery. I am sure some of Knime example workflows will cover it, which I am going to study them.
One of my favorite starting points for small companies is to use the Excel Write to Template node to create portable dashboards for exploration, reports and budgeting.
Basically I build a table that has all of the necessary pieces and settings for the visualizations and on the fly calculations that I would like to be able to create for a dashboard.
I always write the table twice to every excel file. The first write is the raw data (which I typically hide later in a separate tab). Then I reference that raw data to create unique lists that I can use as user friendly data validation drop down settings in the second “editable table”. The visualizations and calculations are based off the editable table (as well as calculated differences). All calculations need to be dynamic, so it will take some googling to pull off excel wizard level array and match formulas to make it fully dynamic and flexible.
I then do conditional formatting on the live editable table to highlight changes from the original data table. This allows adjustments to be made down to the granular level and see the summarized impact instantly reflected in graphs and calculations.
You can also create a second level of summary calculation adjustments as quick UI tools for calcs.
I regularly use this dashboard approach for portable reporting, scenario testing and budgeting when the underlying table is under 20k rows or so.
You can even pre-filter data for specific users / departments and distribute the excel files to recipients via email using KNIME as well.
1 Like
Dear iCFO
Thank you for the information. Except that it is a lot of information. Do you share this workflow in Knime community? Thank you.
I will see if I can put together something basic later in the week using dummy data that I would be allowed to share.
Thank you very much. I hope it is not too complicated. You are basically communicating with a non data scientist and a beginner “idiot” sad to say
The approach moves the heavy lifting of visualizations and interactive calculations over to excel which provides the added benefit of portability. This shortens the learning curve and build time in KNIME (as those final steps tend to require more complex component skills), and leverages excel skills. Unfortunately my dynamic excel calculations will likely be considered complicated…
The approach is still viable if you build “one off” spreadsheets with simpler formulas, but you won’t be able to apply that templates universally as a base for future projects.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.