-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
executable file
·320 lines (294 loc) · 17.7 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
<!DOCTYPE HTML>
<html lang="en"><head><!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-JQSSP7XR3S"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-JQSSP7XR3S');
</script><meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>Santhosh Subramanian</title>
<meta name="author" content="Santhosh Subramanian">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" type="text/css" href="stylesheet.css">
<link rel="icon" type="image/png" href="images/seal_icon.png">
</head>
<body>
<table style="width:100%;max-width:800px;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr style="padding:0px">
<td style="padding:0px">
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr style="padding:0px">
<td style="padding:2.5%;width:63%;vertical-align:middle">
<p style="text-align:center">
<name>Santhosh Subramanian</name>
</p>
<p> I am a recent BA graduate from the <a href="https://math.berkeley.edu"> University of California, Berkeley</a> where I majored in Applied Mathematics with concentrations in financial engineering and machine learning.
I am originally from John's Creek, Georgia and love playing football, basketball, volleyball, chess, and poker! My interests lie in quantitative & mathematical finance,computer vision & deep learning research, sports analytics, data science, and fundamental investment research.<strong>
I am currently
looking for full time jobs, internships, and research positions.
</strong>
</p>
<p>
Throughout my four years at UC Berkeley, I have worked at many different organizations and even started my own -
<a href="https://www.cbinsights.com/company/diascan"> Diascan</a>. Most recently, I worked at <a href="https://www.mainmgt.com/"> Main Management</a>,
an asset management firm, doing primarily quant research. I've also previously worked at
<a href="https://www.kiwibot.com/"> Kiwi Robotics</a>, <a href=https://www.crunchbase.com/organization/gofind-ai> GoFind</a>, and <a href=https://asuc.org/>
The Associated Students Organization of UC Berkeley</a>. On campus, I was involved in Theta Delta Chi and the Quantitative Investing Club @ Berkeley.
</p>
<p style="text-align:center">
<a href="mailto:santhosh97@berkeley.edu">Email</a>  / 
<a href="data/resume.pdf">Resume</a>  / 
<a href="https://www.linkedin.com/in/santhosh97">LinkedIn</a>  / 
<a href="https://github.com/Santhosh97">Github</a>
</p>
</td>
<td style="padding:2.5%;width:40%;max-width:40%">
<a href="images/santhosh_2.png"><img style="width:100%;max-width:100%" alt="profile photo" src="images/santhosh_2.png" class="hoverZoomLink"></a>
</td>
</tr>
</tbody></table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:20px;width:100%;vertical-align:middle">
<heading>Relevant Coursework and Skills </heading>
<p>
<i>Math:</i>   Multivariable Calculus, Linear Algebra & Differential Equations, Optimization, Proof Based Linear Algebra, Real Analysis, Numerical Analysis, Complex Analysis, Honors Abstract Algebra
</p>
<p><i>Computer Science:</i>   Discrete Math & Probability, Structure & Interpretation of Computer Programs, Data Structures, Data Science & Machine Learning</p>
<p>
<i>Financial Engineering (Graduate Level):</i>   Financial Engineering I - Financial Mathematics, Financial Engineering II - High Frequency Trading, Portfolio Theory & Optimization
</p>
</td>
</tr>
</tbody></table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:20px;width:100%;vertical-align:middle">
<heading>Work Experience</heading>
<p>
I have worked at multiple organizations, primarily working with machine learning and data science.
</p>
</td>
</tr>
</tbody></table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<td style="padding:20px;width:25%;vertical-align:middle"><img src="images/main1.jpg"></td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://www.mainmgt.com/">
<papertitle>Main Management, LLC</papertitle>
</a>
<br>
<strong>Quantitative Research Intern</strong>
<br>
<br>
<em>San Francisco, California</em>
<br>
<em>February 2019 to February 2020</em>
<br>
<p></p>
<p>I primarily built prediction
models using alternative data sources to generate signals for when to buy and sell exchange traded funds.
I also developed a tool which scraped historical ETF data from FactSet and Yahoo Finance and
exported elegant reports with various different information such as metrics, valuations, and ratios.
This tool reduced firm's ETF report creation from around two hours to less than thirty seconds.
<i> Utilized Python, PyTorch, SQL, Sci-Kit Learn, Pandas, Seaborn, Matplotlib. </i>
</p>
</td>
<tr onmouseout="nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:25%;vertical-align:middle"><img src="images/gofind.png"></td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://www.crunchbase.com/organization/gofind-ai">
<papertitle>gofind.ai</papertitle>
</a>
<br>
<strong>Machine Learning and Product Intern</strong>
<br>
<br>
<em>Berkeley, California</em>
<br>
<em>Jan 2018 to May 2018</em>
<br>
<p></p>
<p> At gofind.ai, I helped develop their license plate detection application. The first portion of my work was building
the detection model, and the second part was building the character recognition script. The detection model was built using YOLOV2 in tensorflow
and would detect bounding boxes of license plates in images and videos of cars. The character recognition script would then take in
the license plates found in the detection model and output a text file with the license plate numbers. I also solely
worked on the design and wireframing of the application.
<i> Utilized Python, YOLO, TensorFlow, Sci-Kit Learn, OCR, Pandas, OpenCV. </i>
</p>
</td>
</tr>
<tr onmouseout="nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:25%;vertical-align:middle"><img src="images/kiwi1.png"></td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://www.kiwibot.com/">
<papertitle>Kiwi Robotics</papertitle>
</a>
<br>
<strong>Autonomous Driving Research Intern</strong>
<br>
<br>
<em>Berkeley, California</em>
<br>
<em>May 2018 to August 2018</em>
<br>
<p></p>
<p> I collaborated with a development team of four to collect, filter, transform, and analyze over twenty hours of
streamed video data from Kiwi Robots in order to help the robots navigate crosswalks autonomously. My project was to
predict steering wheel and throttle of the delivery robots by implementing the Stanford paper, <a href = 'http://cs231n.stanford.edu/reports/2017/pdfs/626.pdf'> <i>Self-Driving Car Steering Angle Prediction Based on Image Recognition</i></a>,
on the data my team collected. Following the paper, I built two different models in PyTorch. The first model comprised of 3D convolutional layers followed by LSTM recurrent layers. The second model used transfer learning with 2D convolutional
layers on a pre-trained model where I only retrained the fully connected layers.
<i> Utilized Python, Pytorch, Convolutional Neural Networks, LSTM, Sci-Kit Learn, Pandas, OpenCV. </i>
</p>
</td>
</tr>
<tr onmouseout="nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:25%;vertical-align:middle"><img src="images/lime.jpeg"></td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://www.li.me/">
<papertitle>Lime</papertitle>
</a>
<br>
<strong>Data Science Consultant</strong>
<br>
<br>
<em>San Francisco, California</em>
<br>
<em>May 2018 to June 2018</em>
<br>
<p></p>
<p> I worked with the Head of Product @ Lime, Colin McMahon, on a one month long project. The project was to help Lime precisly identify places in San Francisco where tourist activity was high. I
used the Flickr API in order to pull around fifty thousand different location points (latitude and longitude) of images specifically taken from tourists in San Francisco. I then clustered the location points
on a virtual map to help Lime choose which areas to expand too and increase bike capacity. You can check out the clustering <a href="\data\clustering.png">here</a>!<i> Utilized Python, Pandas, OpenCV, Sci-Kit Learn. </i>
</p>
</td>
</tr>
<tr onmouseout="nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:25%;vertical-align:middle"><img src="images/diascan.jpg"></td>
<td style="padding:20px;width:75%;vertical-align:middle">
<a href="https://www.cbinsights.com/company/diascan">
<papertitle>Diascan</papertitle>
</a>
<br>
<strong>Co-Founder and Chief Technology Officer</strong>
<br>
<br>
<em>Atlanta, Georgia</em>
<br>
<em>June 2015 to July 2019</em>
<br>
<p></p>
<p> I started Diascan in the summer of 2015. Diascan was a software that used deep learning and computer vision to help radiologists more accurately
detect and diagnose lung cancer from medical imagery. As the technology lead, I built the machine learning pipeline. The pipeline had two parts. The first
part was a SSD object detection model that would return bounding boxes of tumors found in CT scans. The second part of the pipeline was a
classification model that would classify whether or not the tumor found in the first part of the pipeline was benign or malignant.
<i> Utilized Python, Pytorch, Convolutional Neural Networks, LSTM, Sci-Kit Learn, Pandas, OpenCV. </i>
<br>
<br> Diascan was part of <a href="www.cyberlaunch.vc"> Cyberlaunch</a>, a machine learning accelerator. We also raised a seed round from an Atlanta entreprenuer, Chris Klaus,
and have been featured in multiple news outlets such as TechCrunch and Atlanta Magazine.
</p>
</td>
</tr>
</tbody></table>
<table width="100%" align="center" border="0" cellspacing="0" cellpadding="20"><tbody>
<tr>
<td>
<heading>Projects</heading>
</td>
</tr>
</tbody></table>
<table width="100%" align="center" border="0" cellpadding="20"><tbody>
<tr onmouseout"nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:75%;vertical-align:middle">
<strong>A Comparative Analysis of Machine Learning Models for Breast Cancer Classification</strong>
<br>
<em>CS 198 - Machine Learning and Data Science (Final Project)</em>
<br>
<a href="data/breast.pdf">Paper</a>  / 
<a href="data/breast_code.pdf">Code</a>  
<br>
<p></p>
<p> Using data from the Breast Cancer (Diagnostic) Data set, we built
machine learning models to classify breast mass instances as benign or malignant.
We used five machine learning models - SVM, Random Forest, Logistic Regression, Decision Trees,
and a Neural Network. For each model, we incorporated hyper-parameter tuning and cross validation.
</p>
</td>
</tr>
<tr onmouseout"nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:75%;vertical-align:middle">
<strong>Q Learning for Optimal Trade Execution</strong>
<br>
<em>IEOR 222 - Financial Engineering II (Final Project)</em>
<br>
<a href="data/ieor.pdf">Paper</a>  / 
<a href="data/IEOR_Project.ipynb">Code</a>  
<br>
<p></p>
<p> Optimal Order Execution has been a long standing problem in financial institutions. There are many ways to tackle this issue; however, the backbone
of all the solutions was a strategy laid out by Dimitris Berstsimas and Andrew Lo in the paper <a href="https://alo.mit.edu/wp-content/uploads/2015/06/Optimal-Control-of-Execution-Costs.pdf"> <i>Optimal control of execution costs</i></a>.
The goal of the final project was to use reinforcement learning, specifically Q Learning, in order to learn the optimal order execution strategy under various price impact models that was laid out in Bertimas and Lo's paper.
</p>
</td>
</tr>
</tr>
<tr onmouseout"nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:75%;vertical-align:middle">
<strong>Cal Baseball: Clubhouse Analytics</strong>
<br>
<em>IEOR 185 - Startup Challenge</em>
<br>
<a href="data/clubhouse-demo.mov">Video</a>  / 
<a href="data/clubhouse.pdf">Presentation</a>  
<br>
<p></p>
<p> Worked with the Cal Baseball team to develop an easy to use interface for coaches and players to view and analyze various different sabermetrics (baseball statistics). I was in charge of data engineering and analysis. I
also developed statiscal applications that the team could use such as lineup optimization and in-depth analytical scouting reports.
</p>
</td>
</tr>
</tr>
<tr onmouseout"nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:75%;vertical-align:middle">
<strong> Personal Investing: Stock Pitches</strong>
<br>
<a href="data/netease.pdf">NetEase Long Thesis</a>  / 
<a href="data/nvidia.pdf">Nvidia Short Thesis</a>  
<br>
<p></p>
<p> I have been very interested in investement research. I have written couple of investment pitches for positions in my portfolio. Specifcally, I wrote a long thesis on NetEase and a short thesis on Nvidia. On another note, since January 2020,
my portfolio has beaten the S&P 500 benchmark index by 11%. Portfolio has returned 26.64%, whereas S&P 500 has returned only 15.16%.
</p>
</td>
</tr>
</tr>
<tr onmouseout"nerd_stop()" onmouseover="nerd_start()">
<td style="padding:20px;width:75%;vertical-align:middle">
<strong> Instacart APM Challenge</strong>
<br>
<a href="data/instacart_report.pdf">Report</a>  / 
<a href="data/instacart_presentation.pdf">Feature Presentation</a>  
<br>
<p></p>
<p> Material I wrote and designed for the Instacart APM challenge.
</p>
</td>
</tr>
</tr>
</tbody></table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:50px">
<br>
<p style="text-align:center;font-size:small;">
Design and source code from <a href="https://github.com/jonbarron/jonbarron_website">John Barron's website</a>.
</p>
</td>
</tr>
</tbody></table>
</td>
</tr>
</table>
</body>
</html>