Now you can Instantly Chat with Himalaya!
About Me
I have implemented my project from start to end Using python. I have working experience in python, machine learning, NLP, Django, rest-API, git, MySQL, MongoDB, PostgreSQL. I have worked with the recommendation systems, rest API for the android sy...
Show MoreSkills
Portfolio Projects
Description
Description: Here I have to build user persona according their behavior the platform. According their behavior, I have to push notification, advertise, Real estate property, blog, video etc.
Description
Description: Here I have to build a system where it will recommend user according their search History, according their post, comment, hashtags etc etc. Here I have built a total word relation Mapping. Their I am scraping total 110 websites for latest news update and saving that news In S3 , and later showing those news using the recommendation system according user interest. All of recommendation return back to to user using api endpoint. Role: Lead Developer Technology: python, collaborative algorithm, svr, knn, word2vec, karnel approximation, git, request, Web scraping, s3, sqs, sns, ec2, csv , Django, pandas, numpy, scikit learn, machine Leaning language. Database: Mongo DB Website: https://wenaturalists.com/
Show More Show LessDescription
Description: It’s a desktop software where kids can learn to write “a,b,c,d” which is made using tkinter and matplotlib as main component. There every letter is a dynamic graph using matplotlib. Website: https://www.startwriteindia.com/ Role: Developer Technology: python, matplotlib, tkinter, git, numpy , pandas Database: Sqlite3 Language: Python
Show More Show LessDescription
Description: Creating a completely new cryptocurrency called arcturus. Developing the website and building the coin was my responsibility. I used ethereum ico base to create this coin. Website: www.arcturus.com Role: Lead Developer Database: Mongo dB, MySQL Technology: Block chain technology, aws ec2, blockchain, Django,git Language: Python, html, css , javascript, Solidity, Natural language processing
Show More Show LessDescription
Description: Here I have to build a system where it will Extract the text from image, Translating the Text into other language, rewrite on the image. I am using yolo algorithm with cnn to detect the Text, Google translate api for translate the text, Image Inpainting for removing and repairing the image. Role: Lead Developer Technology: python, yolo algorithm,TensorFlow, opencv, cnn, Django, git Database: MySQL
Show More Show LessDescription
Description: It will scan a website and find out those vulnerabilities which can affect security of that website. At first, we have to bypass the load balancer of a website. So, I have to use change user agent randomly. I used various type of payload to detect those vulnerabilities. Role: Developer Tools: Accunetix, apache airflow, git. Technology: Mechanize, owasp top 10 vulnerability architecture, Json, xml, python,sql Database: Mongo dB, MySQL
Show More Show LessDescription
Description: predict clusters of customers by store location. knowing the set of customers that behave evenly will help them target their product promotions accordingly. In this project data wrangling is the main problem. There is to much error in the data. But after filling the data in particular way I am able to solve the problem. Next problem was the predicting. I have been used pca over gmm and k-means algorithm to solve the problem.
Description
It’s a Python package which is update by me from previous bezier algorithm package. I have been implanted linear interpolation method on the previous algorithm. Now this package will return points and axes plot as per your requirement. It can ignore some angle like you don’t want a curve under 30 degrees.
Show More Show LessDescription
It’s a serverside tool. It’s can work in real time and also can analysis server log file to identify icmp smurf attack. It’s totally made in python with help of keras and tesorflow. I have been created 6 neurons and the accuracy was approx 92% where for this same random forest algorithm was returned accuracy of 78% with 10000 n_jobs and 1000 nodes with lot more time complexity.
Show More Show Less