About
Articles by Shubham
Activity
-
How to 𝐛𝐮𝐢𝐥𝐝 𝐚𝐧𝐝 𝐬𝐜𝐚𝐥𝐞 𝐚𝐠𝐞𝐧𝐭𝐢𝐜 𝐨𝐫𝐠𝐚𝐧𝐢𝐳𝐚𝐭𝐢𝐨𝐧𝐬 is always my favorite question. At Google, I made many observations…
How to 𝐛𝐮𝐢𝐥𝐝 𝐚𝐧𝐝 𝐬𝐜𝐚𝐥𝐞 𝐚𝐠𝐞𝐧𝐭𝐢𝐜 𝐨𝐫𝐠𝐚𝐧𝐢𝐳𝐚𝐭𝐢𝐨𝐧𝐬 is always my favorite question. At Google, I made many observations…
Liked by Shubham Saboo
-
Voice dictation AI app for your Desktop. Hold one key. Speak. And the text appears perfectly. Just added to Awesome LLM Apps GitHub Repo. 100%…
Voice dictation AI app for your Desktop. Hold one key. Speak. And the text appears perfectly. Just added to Awesome LLM Apps GitHub Repo. 100%…
Liked by Shubham Saboo
Experience & Education
Licenses & Certifications
Volunteer Experience
-
Technical Reviewer
O'Reilly Media
- Present 5 years
Science and Technology
Reviewed the following technical books and reports:
> Practical MLOps: Operationalizing Machine Learning Models
> Report: Enterprise AIOps -
Community Volunteer
Open Data Science Conference (ODSC)
- 1 month
Science and Technology
Worked with ODSC team to organize one of the largest data science conferences in APAC.
> I was responsible for speaker logistics and support. Moderated multiple live sessions by distinguished professors, industry experts, and global leaders in data science.
> I also played a key role in creating a conducive community environment for the attendees to easily interact with the speakers and other attendees. -
Freelance Journalist
India Today
- 1 year
Education
Here are some of my most popular pieces:
> Explained: Why this is the right time for engineering students to study in India than in foreign university? (https://www.indiatoday.in/education-today/featurephilia/story/explained-why-this-is-the-right-time-for-engineering-students-to-study-in-india-than-in-foreign-university-1763484-2021-01-28) -
Technical Writer
Towards AI
- Present 6 years 7 months
Science and Technology
Write technical blogs based on reader-centered pedagogy in the area of Artificial Intelligence, Machine Learning, and Data Science.
> Find my blogs at shubhamsaboo111.medium.com -
Technical Writer
Analytics Vidhya
- Present 6 years 6 months
Education
Write technical blogs based on reader-centered pedagogy in the area of Artificial Intelligence, Machine Learning, and Data Science.
> Find my blogs at shubhamsaboo111.medium.com
Courses
-
AI at Scale
-
-
Data Driven Design Thinking
-
-
Data Visualization
-
-
Data-X (In collaboration with University of California, Berkeley)
-
-
Decision, Risk and Optimization
-
-
Design Thinking
-
-
Designing Our Lives
-
-
Engineering Ethics
-
-
Graph Theory and Social Network Analysis
-
-
Indian Society and Culture
-
-
Law, Society and Technology
-
-
Machine Learning for Genomics
-
-
Machine Learning for IoT
-
-
Natural Language Processing
-
-
Product Design and Management
-
-
System Thinking
-
-
The Role of Markets
-
Projects
-
Financial Market Time Series Prediction with Recurrent Neural Network
In this project I have used Long short term memory(LSTM) networks, a subclass of recurrent neural networks using keras with tensorflow (backend), to predict stock prices of Google. The network outperformed kalman filter, predicting more of high-frequency fluctuations in stock price.
-
Counterfeit Currency Detection Using Image Processing
In this project, I have used the Bank Authentication Dataset from the UCI repository for training the neural network model. Using wavelet transformation of an image it is predicted whether or not a Bank Note is authentic.
-
Credit Risk Assesment Model
In this project, I have used data from LendingClub.com. A machine learning model based on Support Vector Machine is created that will evaluate the potential risk posed by lending money to consumers and predict people with a high probability of paying back.
-
Multi-Class Semantic Text Classifier (NLP)
-
Created a multi-class text classifier by fine-tuning the Google's BERT(Bi-directional Encoder Representation from Transformer) to automatically categorize the the stack-overflow queries into respective technologies.
BERT creates contextual pair wise sentence embedding to encode the input text provided in the form of natural language (i.e English) and uses transformer based network architecture that is pre-trained on Wikipedia and large corpus of books for semantically classifying text…Created a multi-class text classifier by fine-tuning the Google's BERT(Bi-directional Encoder Representation from Transformer) to automatically categorize the the stack-overflow queries into respective technologies.
BERT creates contextual pair wise sentence embedding to encode the input text provided in the form of natural language (i.e English) and uses transformer based network architecture that is pre-trained on Wikipedia and large corpus of books for semantically classifying text into multiple categories.
The data used is from stack-overflow which consist of user queries as text and respective technologies as labels. The size of data used is of 20,000 samples spread into 20 different categories (i.e technologies). -
Human Pose Based Gait Recognition
-
Created a LSTM based network to recognize human gait patterns based on input key points extracted from the videos using the OpenPose library to create an efficient and robust surveillance system.
Honors & Awards
-
Microsoft Most Valuable Professional (MVP)
Microsoft
Issued in recognition of your exceptional leadership and outstanding contribution in the field of AI.
-
International Informatics Olympiad
-
Secured 341 World rank among 200k+ participants in International Informatics Olympiad, 2011
Recommendations received
8 people have recommended Shubham
Join now to viewOther similar profiles
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content