Suhas's Portfolio

Work experience
July 2018 - Present
Senior Software Engineer

Stripe, Seattle, WA.

Designed & implemented a service to shield Stripe’s systems from millions of spurious requests/day and L7 volumetric DDoS attacks at the edge of Stripe’s network

Built an end to end system to perform reachability and latency monitoring for Stripe API to ensure Stripe’s global availability.

Designed & implemented a second CDN for Stripe to build resilience against single provider outages and unlock new static asset delivery capabilities.

Designed & implemented a platform to perform sanction screening and consent management for Stripe’s consumer products.

Founding engineer on multiple teams; helped establish strong norms for engineering & operational excellence and building the culture of user’s first for delivering solutions.
Feb 2016 - June 2018
Senior Member of Technical Staff

Salesforce.com, San Francisco, CA.

Designed and developed scalable and distributed data pipelines for Salesforce IoT to perform data transformations.

Developed the Ingestion pipeline for high throughput events using Kafka and Azure Event Hubs.

Developed Output Connectors running in Kubernetes for the processed data to be utilized from Salesforce IoT to the outside world.
Sept 2015 - Oct 2015
Insight Data Engineering Fellow

Insight Data Science, NYC

Built a Big data pipeline to recommend parking spots in real time based on the parking sensor data of San Francisco city

Code demo

Technologies used: Ingestion: Kafka, Storage: Hadoop(hdfs) Batch Processing: Apache Spark, Stream Processing: Spark Streaming, Database: Cassandra, Web server: Python - Flask
May 2015 - Aug 2015
Software Engineer Intern

Salesforce.com, San Francisco, CA.

Building a data management tool on the Service cloud using Force.com REST and SOAP API Web services.

Technology Used: Python, Java.
August 2014 - Current
Research Assistant, College of Education, University of Illinois at Urbana-Champaign, IL.

Full stack web developer

Developed an online Degree Qualifications Profile(DQP) Assignment Library, which helps professors review, submit and share assignments with others across universities worldwide.

Technology Used: MEAN Stack (MongoDB, ExpressJS, AngularJS and NodeJS)
July 2012 - June 2014
INTEL INDIA

Software Engineer, Bangalore, IN
- Worked as Lead Front End Web DeveloperBuilt responsive, large scale web application which enhanced productivity and user experience for various scheduling applications across Intel.
- Enterprise Issue Management (EIM) - Developed a Customer facing Web application for Intel Warranted Product Issue resolutions.
June 2011 - August 2011
Thomson Reuters

Software Developer - Intern in Bangalore, IN

Reuters Market Light (RML) Implemented Collaborative filtering algorithm to identify agrarian services like Crop Prices, Weather Forecast, Cash Crops etc. and send them to the subscribed farmers over mobile using SMS gateways

Technology Used: Java, J2EE with Struts 2.0 Framework

PROJECTS
Real-Time Twitter streaming analysis of Top N Hashtags using Apache Storm.

Technology Used: Java • Redis- advanced Key-Value cache • D3.js • Flask Web server • Vagrant
Meta-crawler: Review Analyzer.

Developed web portal to aggregate, index, summarize, analyze and visualize the sentiment of user reviews of smart phones in real-time.

Technology Used: AngularJS • D3.js • ElasticSearch • Python - Flask • Heroku
Handwritten Digit Classification using Perceptron and Naive Bayes Classifier. [Java]
Built a real-time emoticon recommendation engine using Socket.io.
Data mining on the Social Web - Twitter and Facebook.

Technology Used: Facebook graph API • Twitter package for python • iPython notebook

URL: http://suhashm.github.io/Social_Web_Mining/
Pacman game solving agent using A* search algorithm. [ Java ]
Implemented Naive Bayes with AdaBoost ensemble method to predict whether an auction car is a good buy-Kaggle competition. [Java]
Developed a Reinforcement learning based solution to identify optimal policy of a Markov decision process. [Java]
Implemented Recursive Backtracking search to solve a course scheduling Constraint Satisfaction problem. [ Java ]
Implemented the Apriori algorithm for frequent pattern mining to identify representative phrases on a corpus of research papers. [ Java ]
Tech Pandora - MEAN Stack Application

Full Stack web application written completely in Javascript using MEAN Stack - MongoDB, ExpressJS, AngularJS and NodeJS

URL: http://suhashm.github.io/MEAN-Stack/
Node Authentication

Implemented Local SignIn and SignUP, Facebook, Twitter and Google+ authentication using Passportauthentication middleware for NodeJS

Technology Used: NodeJS, ExpressJS, PassportJS, MongoDB

URL: http://suhashm.github.io/node-authentication/
Optical Character Recognition - Android Application

Developed an Android application in which neural networks are trained using Machine Learning Algorithms like Kohonen self-organizing maps and Hilditch Image Processing algorithm to recognize large data sets of text from an image and the subsequent characters are optically recognized and read out using native text-speech engine.
e-Scrap Mobile Application at Intel Code-A-Thon

Developed a mobile application to track the e-scraps from Intel Labs

Technology Used: HTML5 and jQuery Mobile for front end and MySQL for backend and used PaaS (Platform as Service) to physically land the app in Intel Enterprise Private Cloud.
Single Pass Assembler for 8086 Microprocessor

Implemented an assembler using a lex tool named flex and devised our own algorithm to handle forward references while parsing

Used Assembly language programming to build a dictionary with instructions and generate the assembly code in a single pass.