Rishan Shah

Data Scientist and Tech Enthusiast

Warning this website is always experiencing changes

Welcome!

Thanks for stopping by! Here you will find some of the things I have made and been involved in. I love all things Data Science and also have a deep interest in Cosmology and Theoretical Physics

Perumal Kumaresa

(Data Science Manager)

"I had the opportunity to work closely with Rishan on multiple initiatives and projects at Data Reply. Rishan is one of the hard and technically sound people, he is self-motivated and thrives to learn and share knowledge where ever he can. I am sure his continuous hunger to learn and innovative thinking by applying AI & Machine learning to solve real-life & business problems will be valuable for any company he is part of. I highly recommend Rishan Shah for his machine learning & analytical skills alongside his excellent presentation and delivery ability."

Stefan Blaginov

(Blockchain Engineer)

"I've rarely met someone that can grasp new concepts as quickly as Rishan. This along with his well-developed mathematical intuition, made working with him a pleasure."

Jonathan Orritt

(Data Scientist)

"Rishan was a core asset to the Payments Data Lab at Santander. He demonstrated a broad range of core Data Science skills and excelled within his role. His ability to apply advanced machine learning to applicable business problems was invaluable."

Project Spotlight

Got really into one of my personal projects into understanding how reinforcement learning works and decided to give it a shot. Ended up creating a "Deep Q Reinforcement Learning model" that optimises price based on a cost sensitive poisson arrival distribution

Something Interesting

Article: Network Analysis Watford's Vehicular Road Network

A really interesting and exiting project as part of my masters programme whereby I analyse vehicular road networks in Watford and apply Network Science Techniques to establish vulnerabilities

View

Article: Latent Semantic Indexing with Weighted Cosine in Low Rank Appoximation

Here is an article about something I came across but wasn't documented anywhere. I attempt to approximate matches of sentences with an importance on words in a highly compressed storable form

View

Oshwal Calendar

I decided to put some of my new knowledge to use and built a NodeJS calender subscription API for my community events which my family follows. I deployed this into heroku. Unfortunately, this doesn't work for Android devices since they dont support the webcal protocol.

Subscribe

BSc. Mathematics, Operational Research and Statistics from University of Cardiff.

MSc. Data Science from University of Exeter.

Reply

Dec. 2021 - Present

Work with time series forecasting.
Work with Association Rules.
Work with Graph Machine Learning.
Work in FSI and Real Money Gaming industries.
AWS Machine Learning - Speciality accreditation.
AWS Certified Cloud Practitioner accreditation.
Dataiku ML Practitioner accreditation.

TUI

May. 2019 - Dec. 2021

Engaged with NLP uses cases.
Work with Reinforcement Learning.
Experience with gym, coach & tfagents.
Use of AWS Sagemaker.

Santander

Sep. 2017 - Apr. 2019

Used Big Data technologies such as Hadoop and Spark to clean and enrich data.
Made use of Python libraries such as scikitlearn, pandas, plotly, seaborn, keras and tensorflow to deliver insights.
Worked with teams under "Agile" methodologies.

CACI

May. 2017 - Aug. 2017

Delivering weekly reports to senior managers.
Creation of insights through the use of Adobe Analytics and Google Analytics.
Understanding and exploiting digital marketing strategies that overlayed client websites.

Skills

Python

Pyspark

SQL

Data Science

Machine Learning

Deep Learning

Git

In Conclusion

Rapidly progressing in my knowledge of the new cutting edge techniques in data science. Always willing to sponge up whatever knowledge more experienced people are open to offer. Lastly, I live by the following two wisdoms:

Data Science is not Rocket Science, you only need to know which tools and technologies to use
Data Science is not for the well-learned but the data hungry