ocinternship

[ocinternship]

1/21/14 - RRrrr, I'm a Pirate!

Today, I put the finishing touches on the functions I was working on last week. Usually I will have a few runtime errors or some faulty coding that prevents achieving accurate results. I believe I fixed all problems related to the helper functions I have been building in R to help format datasheets for easier data analysis. After I finished up my functions, I went back to playing around with datasheets in R. I am trying to find some correlation involving time and contacts made.

After lunch, I looked at Vashist’s analysis work done in MATLAB. At 2 o’clock, we had a stand-up meeting where we all discussed what we had been working on, as well as company questions/goals/etc. This was one of the more intense stand-ups I’ve been able to be apart of because they have a client who wants to be able to use the product sometime before early February. Thus, everyone has been working hard to make sure that they will be able to meet the deadline and deliver exceptional software. Hopefully Vashist will be able to start coming in on Wednesday’s so that I will be able to discuss analytics with someone much more experienced. Under his tutelage, perhaps I can do something more substantial with all the R knowledge I’ve been gathering.

Here’s a wise quote from Brian: “if you want to create a successful startup, just design a product that will melt people’s faces off.”

Working Hours: 7.5

[ocinternship]

1/14/14 - Sick

Did not go to internship.

[ocinternship]

1/7/14 - R Function Building

I’ve almost finished programming my prior Excel analytics into a code in R. I’ve started to play around with R and have begun creating my own functions. I have many hypotheses and ideas that I want to try coding in R. I’m encountering small syntactical problems here and there because the way I taught myself R was quite crude. I skipped a few fundamentals and jumped into the harder built-in functions. It becomes quite frustrating when I can’t proceed with my code because I never learned the more basic functions. I’m starting to take a closer look at a R file with interesting functions on predictive model building. I plan on incorporating some of those functions next week into my code.

Here’s a wise quote from Dmitry: “now that you’re in college, it’s the perfect time for you to ask for a car, dye your hair, and see how bad your grades can dip before getting kicked out of college.”

Working Hours: 7.5

[ocinternship]

12/17/14 - Even More R Programming

Nothing too extraordinary happened. I am continuing to transfer my pervious analytics done within Excel into a similar R code.  Today I was allowed to sit in on a conference call. Although the call never actually occurred, I was able to learn more about LinkedIn, networking, and how to formally conduct a conference call with a potential client. Right before I left the office I was given the opportunity to screw in the bulb for one of the lights for the menorah.

Here’s a wise quote from Roy: “a snow day is dependent on whether you choose to do your homework or not.”

Working Hours: 7.5

[ocinternship]

12/10/14 - More R Programming

Since I will no longer be doing analytics within Excel, I can happily say goodbye to those dreadful loading times. Analytics done within RStudio is much faster than those done in Excel. The only difficulty I am currently facing is that I now have to learn a whole new language. Excel required very little programming knowledge, but since R is an actual programming language itself, I have to dedicate time in learning basic syntax. Luckily, I know a few basic programming languages, so learning R was not a totally new experience. Today I primarily read, watched, imitated, and learned how to properly do basic analytics in R. I am currently unable to do regression analysis in R, but that is the goal for next week.

  In order to celebrate the upcoming holiday season, we had lunch at the Union Square Café. It was really fancy and the food was top notch. Its small social gatherings like this that really make me interested in perhaps one day joining or creating my own startup. The small work environment is very preferable to me.

Working Hours 7.5

[ocinternship]

12/3/14 – Programming in R

Today I reviewed a sample code Roy had made in the programming language R. In RStudio, I ran the code line-by-line, read documentation when necessary, and commented on the function of each few lines of code. The code is incomplete, so I worked on fixing and adding to the code. I believe that programming in R is more efficient than using Excel for data analysis even though I may have to spend a few more weeks learning the language. Although Excel is very user friendly, RStudio runs R code extremely quickly compared to Excel. Also, R makes it easy to understand what work I was or am working on. R also helps make project organization much more simplistic. Once I am able to transfer the work I had done on Excel as an R code, Brian will be able to later transfer my work onto the actual software through Python.

Working Hours: 7.5

[ocinternship]

11/26/14 - Logistic Regression

Because Thanksgiving is tomorrow, many of my coworkers were either not in the office or were leaving early. Roy had told me that he had met with a friend and wanted to adopt logistic regression analysis. Although similar to linear regression, logistic regression is used to predict a binary response from a binary predictor. It is very useful with setting up qualitative response models, which is beneficial to me because some of the data I work with is qualitative. Roy sent me some assignments to look at so that I could understand how I might approach creating logit models.

Working Hours: 6

[ocinternship]

11/19/14 – Optimization

I found an open source solver tool for Excel that was able to handle the optimization I wanted. The percentage increase value I received from using the open source solver seemed a little too far-fetched even though there were very minimal constraints. I looked through all my older files and cleaned up many of the sloppy coding. I also got rid of duplicates that I somehow failed to discover earlier. Much of the work I completed today was simply cleaning up small errors and organizing my files.

Working Hours: 7.5

[ocinternship]

11/12/14 - Excel Solver Tool

Today Roy and I discussed about how I could be using the current data I have collected of sales people to find the largest amount of sales that could be made based on reassigning calls to different sales people using certain constraints. Essentially what Excel’s solver tool does is finds a solution that satisfies all constraints and maximizes or minimizes an objective cell value. Although Excel’s solver tool is powerful and has many applications, it can only work with a little over 200 cells.  I didn’t realize this until the end of the day when I had finished setting all the constraints I wanted and tried to optimize sales. Roy and I will both look for some replacement for the solver tool that can handle a large amount of different cells. 

Working Hours 7.5

[ocinternship]

11/5/14 - Picking Up the Pace

I arrived to the office, and just like in the previous weeks, I went on the infamous donut run with Brian. Dunkin’ Donuts recently released a croissant donut—it’s basically a thicker glazed donut—and Brian generously bought one for all the Opticlose employees, including myself. We talked about the usual weekly company updates, colleges, and other miscellaneous things.

Back at the office, Abel helped me setup a Slack account under the Opticlose domain. Slack is a platform for effective team communication that utilizes concepts of chatrooms, file sharing, and instant messaging. Because everyone is typically working on their own individual tasks, group emails were previously used as a means of keeping everyone informed. However, emails quickly and easily fill inboxes, so Opticlose shifted towards communication software for easier accessibility in determining the overall progress of the company.

I continued working on analyzing call data in search of interesting correlations involving time period between calls. Following Roy’s advice from last week, I worked on creating graphs and implementing them in a PowerPoint. I’ve come to learn that visuals are very helpful in demonstrating data and that creating PowerPoints of my findings helps in presenting and explaining those results. I’m also really beginning to enjoy the environment that a startup provides because I have the freedom to work where I want, it’s easy to communicate with employees because they are inches away from me, and it’s very friendly.

Working Hours: 7.5

[ocinternship]

10/29/14 - Slow Day

Sadly, no trips to Dunkin’ Donuts were made on this day. Roy and Abel had gone to a conference that week in which they presented the product in design. The two also presented their presentation for the rest of the Opticlose employees. In order to gain more early exposure for the product, Roy has decided to start beginning to spread the Opticlose name.

Today, I started to analyze data with the main focus on time between calls. So far, I have primarily worked on analyzing diminishing returns of calls, but I have not yet incorporated the important factor: time. I mainly cleaned up spreadsheets for specific data columns I planned to use and created some basic formulas to get started on my analytics. Roy suggested that I begin recording my results in a PowerPoint so that I would have a place to store my findings. Also, for future events, they may plan to present some of my results. My results have not yet been verified, but I am trying my best to prevent any errors in incorrectly drawing inferences or creating incorrect equations.

Working Hours: 7.5

[ocinternship]

10/22/14 - Halcyon Days

On the trip to Dunkin’ Donuts, Brian lectured me about computer clusters and how Opticlose would soon begin to use DigitalOcean for their own cluster system. Strangely enough, I had been learning about bus systems and its relations to computer clusters in my electrical engineering class that week. Brian is extremely informative when it comes to computer science, and those Dunkin’ Donuts trips continue to fuel my interest in computer science.

When I arrived back at the office, I spoke to Roy about what I had done the previous week when he was absent. I showed him the spreadsheets I had been working on and any hypotheses I was coming up with. I enjoy my discussions with Roy as he typically has a strong understanding of what direction certain data should be taken for it to have actual use. I began working on analyzing the diminishing returns of each call made and arrived at some interesting results. Although neither Roy nor I predicted such an outcome, we decided that those results may prove useful in the future. I also created a spreadsheet to track an individual person’s ability to close calls in different categories. I was able to determine a few patterns and made note of them within a conclusion tab. My day concluded like much of the other weeks, but I was happy with the improvement I was making.

Working Hours: 7.5

[ocinternship]

10/15/14 - Without a Mentor

Roy was going to be out today, but I already knew what I wanted to get done. Before I could begin working on my prediction analysis, Brian invited me to Dunkin’ Donuts. Since donuts and coffee make up the breakfast of champions, there was no way I could refuse. It seems that Brian has become the one to fill me in every week on the status of the company. He also talks to me about his experiences in the film industry, personal life, and projects he is currently working on for Opticlose. He’s a very interesting fellow.

I continued analyzing contact attempts and its correlation with other data categories. I created a spreadsheet to determine the chance of closing a sale at each contact attempt for different categories. I also created a spreadsheet for determining the same exact thing, but I made the range of contact attempts adjustable. My goal is to optimize this data to determine the best contact attempt range to close a sale based on category. I started playing around with heat maps in Excel for a better visual understanding where the highest percentage chances were located.

For lunch, I went to a nearby Chipotle and ordered a bowl to go. Eager to continue working, I took my lunch back to the office. Because my hands were occupied, I simply took a “step backwards” and looked at my analysis spreadsheet as a whole. I was able to see the general direction in where my analysis was heading and jotted down a few notes as to what my expectations were. I realize that sometimes I get too focused on the current issue I am working on that I forget why solving the issue was critical in the first place. From now on, I’ll begin writing down tasks and their purpose to help me conquer this problem.

[ocinternship]

10/8/14 - Making Progress

From now on, it seems that the train and PATH will become my main form of transportation. For the first time, I was able to successfully travel to the office without any mishap or hardship. I arrived at the office at 9 and only Brian was there. Neither of us had eaten breakfast yet, so Brian took me to a nearby Dunkin’ Donuts. On the way and back, we talked about company updates, issues, and programming.

I continued working on analyzing data and looking for trends in the amount of contact attempts necessary for a sale to close successfully. I can’t disclose too many of my findings, but I am making a lot of progress. After reading much of Excel’s documentation on certain formulas, I have been able to manipulate data to represent what I want successfully. I have been improving on using Excel’s total functionality including working with duplicates, tables, organization, and working with multiple sheets. Because the original Excel file with all the data is so large, it’s hard to work on it because of the lag it generates. So, besides simply analyzing, I create spreadsheets with bits and pieces of data from the whole dataset because it doesn’t slow down my laptop. It also makes it easier for me to keep things organized and work efficiently since I don’t always need all the data open.

It was Dmitry’s birthday today, so Roy treated us all to Rosa Mexicano. I met Abel, the man in charge of product and design at Opticlose. He’s not a full-time employee, so I hadn’t met him until today. We all walked back to the office and we talked about colleges. Roy’s goal is to dissuade me from pursuing finance so diligently because he believes that I am still too young to know where life may take me. Surprisingly, I find this somewhat true, as I have started to become fascinated in computer science due to this internship.

Working Hours: 7.5

[ocinternship]

10/1/14 – The Real Data Has Arrived

Last week Brian had suggested that that I take the train instead of the bus because it may be quicker. Not only did I find the train much quicker than the bus, but the train is also much more comfortable. On my walk to the office, a sign that read 14th Street West deceived me. I walked in the opposite direction assuming I would arrive on 14th Street East, but that was clearly not the case. Later when I thought about it, I realized—duh—one has to travel through the west side before one can reach the east side. I still managed to arrive at the office at 9, but nobody else was there yet. Brian is typically at the office very early, but I later found out that he was sick and would not be coming to work. I looked through my notes from last week and worked with the wine data from the previous week.

When Roy arrived, I showed him the analysis I had done on the data set. He then showed me the utilities within the Analysis Toolpak in Excel. We did linear regression analysis on a dependent variable and multiple independent variables, and we discussed the importance of the coefficient of determination as a predictor in statistical models. We also talked about the flaws of linear regression analysis and how we might be able to intuitively or mathematically tackle these issues. It is always beneficial to have a graph of the data in order to make hypotheses based on visual analysis. Because data can have linear, exponential, clustered, and random trends, linear regression analysis is not always suited. However, Roy and I discussed the various methods that can be used to linearize certain data sets for linear regression analysis.

Roy then sent me a 20mb Excel file of the actual data I would be analyzing. From the file size, I could tell that I would be working with thousands upon thousands of data. I went out to go buy lunch while the file was being uploaded to my computer. The great thing about interning in NYC is that there are tons of places to eat. I decided to get a pita from The Hummus & Pita Co. and went back to the office to begin working on the data sheet. There were over 200,000 rows and 50 columns of data. It was quite intimating—maybe because my laptop starting running really slowly—, but I worked on this data sheet for the rest of the day. I made various leads, but I also fell short when I realized my Excel formulas were inaccurate. It’s very easy to lose track of what exactly is going on in an Excel formula, so I started compiling all my formulas in a document and making note of what they do. I left the office at 5:30 and took the 6:20 train back home. On the train, I thought about how I could execute my analysis with less of a margin for error. Even though I get home quite late, I thoroughly enjoy the work and appreciate all the assistance and training I am currently receiving from my coworkers.

Working Hours: 8

[ocinternship]

9/24/14 - The First Day

Roy gave me freedom in choosing my own working hours, so I decided that I would aim to be in the office for at least 7 hours each week. I hopped on the 7:30 bus, slept soundly, and arrived at Port Authority. Having traveled to the company just last week, I correctly navigated through the underground tunnels and subway system to Union Square. From there, I walked to the office and arrived at 9:30. 

Brian was the only one there at the time, so he helped setup my laptop for any future programming work that I may be involved in. I created a GitHub account and familiarized myself with its interface. Then, I installed Xcode and Homebrew. Brian then demonstrated how Homebrew simplifies software installation. We also talked about why open-source software is typically preferred in smaller businesses. Because that type of technology is free to use with little to no licensing restrictions, smaller businesses are more likely to receive funding due to the lack of extra costs in licensing software.

Roy soon came in and debriefed me about the overall goals of the company, their current standings, and the problems they want to overcome. Since today was mainly a learning day for me, Roy sent me links on linear programming, Excel optimization, and simple linear regression. I brushed up on my statistics knowledge and learned how to do simple linear regression analysis. I played around with wine data Roy had sent me in Excel. Since I did not have extensive knowledge of Excel, I asked Roy for assistance when I encountered problems or had questions. I definitely need to improve my knowledge of Excel because it is a powerful tool for data analysis. At 2 o’clock, the company had a ‘checkup’. This was essentially a short session in which we all discussed what we had completed in the morning, what we were working on, and any company-related topics. After I had finished working with the wine data set, I wanted to continue learning more about the programming aspect of the company. Dmitry invited me to the company's repo on GitHub, and Brian showed me the basics of using GitHub—add, commit, and push still resonate in my mind. Growing quite fond of computer science, Dmitry and Brian directed me to some sites I could use to familiarize myself with different programming languages. I definitely cannot wait to come back next week and start working with actual data for the company.

Working Hours: 7

[ocinternship]

The Beginning

On September 17th, I had an interview for an internship at an early-stage startup based in NYC. That was the first time I had traveled to Manhattan alone. It seemed daunting at first, but I was able to successfully navigate myself through the city with the help of the bus, subway, and my savior – my iPhone. So, after having successfully landed the internship, I became ecstatic to embark on this whole new experience. I am currently a senior in high school and having a weekly internship, known as ’senior experience’ in my school, is part of our curriculum. I will be blogging about my experience each week. Without further ado, here is how my senior experience unfolds…