Hi, I'm Prathamesh Saraf.

A
Self-driven, learner-on-the-go, passionate programmer. A curious mind who enjoys solving a complex and challenging real-world problems. Vision is to use technology to make a difference in the lives of people.

About

I am an M.Tech Research post-graduate student in Computational and Data Sciences(CDS), Indian Institute of Science (IISc), Bangalore.

Currently, part of Cloud Systems Lab (CSL) led by Dr J Lakshmi. Interest lies in Systems, Virtualization, Reinforcement Learning, Chatbot development and Machine Learning.

Experience

Chatbot Developer
  • Design, develop, and maintain multilingual text and IVR-based chatbots.
  • Worked extensively on RASA, modified its open-source code to generate automated test cases from current and past conversations of users.
  • Setup analytics services for chatbots that help in generating business leads.
  • Streamlined bot development process.
  • Tools: Python, Flask, Rasa, Azure, Aws, Redis
August 2020 - March 2021 | Bengaluru, India

Projects

quiz app
Atomic Broadcast Protocol

Implemented a distributed token based atomic broadcast protocol as part of Distributed Systems Course.

Accomplishments
  • Tools: Python
  • Implemented a distributed token based atomic broadcast protocol
  • Demonstrated the protocol using 2 broadcast servers and 2 application servers.
  • Each application message had a global unique sequence number, with acknowledgement for each message packet received
quiz app
C BOT

A chatbot to teach C programming.

Accomplishments
  • Tools: Rasa, Flask, Telegram, Azure, StackOverflow Api
  • A chatbot developed using RASA which helps users solve their C programming queries.
  • Integrated with StackOverflow API to resolve complex queries.
  • The bot is deployed on the Azure cloud and can be interfaced using Telegram and AimyBox.
music streaming app
Intelligent Parking Assistant

Deep learning based parking assistant for autonomous cars

Accomplishments
  • Tools: Tensorflow, Python, Pytorch, OpenCv, Image Annoter
  • Yolact based image segmentation model to identify parking slots in Indian conditions
  • Reeds Shepp based car trajectory model to find (semi)optimal way to park the car.
Screenshot of web app
Mitra - Your Friend in Crisis

Created with love a chatbot powered by wit.ai to help out people in COVID pendemic.

Accomplishments
  • Tools: Wit.ai, APIs, HTML, CSS, Bootstrap, Flask, Python
  • A chatbot created using ‘Wit.Ai’ to help people during emergency situations.
  • The bot is deployed on Heroku cloud and can be accessed using Facebook Messenger (Created as a part of Facebook AI Hackathon).
Screenshot of web app
Netify

Telegram bot to give model training updates.

Accomplishments
  • Tools: Telegram Api, Keras, TensorFlow, Flask, Python
  • A telegram bot to send training updates on the go, right on your telegram app.
  • No need to wait long hours just watching your model getting trained.
Screenshot of web app
Mouse Gesture

Controlling computer mouse movements with hand gestures

Accomplishments
  • Tools: OpenCv, PyAutoGui, Python

Skills

Languages

Python
C
C++

Libraries

NumPy
Pandas
OpenCV
scikit-learn
matplotlib

Frameworks

Rasa
Flask
Keras
Django
PyTorch
TensorFlow
OpenAI Gym

Education

Indian Institute of Science(IISc), Bangalore

Karnataka, India

Degree: M.Tech(Res) in Computational and Data Sciences (CDS)
Lab: Cloud Systems Lab

    Relevant Courseworks:

    • System Virtualization
    • Distributed Systems
    • Reinforcement Learning

DKTES Textile and Engineering Institute

Maharashtra, India

Degree: Bachelor of Technology in Computer Science and Engineering
CGPA: 9.24/10

    Relevant Courseworks:

    • Data Structures and Algorithms
    • Operating Systems
    • Computer Networks
    • Database Management Systems
    • Machine Learning

Miscellaneous

Certifications

  • Rasa Certified Chatbot Developer
  • Andrew Ng: Machine Learning Coursera
  • Deeplearning.ai - Neural Networks and Deep Learning
  • Deeplearning.ai - Deeplearning.ai: Improving Deep Neural Networks
  • Crio Student Developer

Achievements

  • Runner up in RIT ML Hackathon.
  • Ranked 5540 in Google Kickstart 2020
  • Valedictorian in throughout Engineering.
  • Ranked 4282 (Team: Warriors) in Google Hashcode 2020
  • Ranked among the top 5% (2821 / 55654) in TCS code vita 2k19
  • Among the top 10 teams in the E-Yantra robotics competition of IITB from 100 teams.
  • The winner among 100’s of contestants of the ML Hackathon on HackerEarth hosted by 100 Days of ML.

Contributions

  • Founding member of ‘Coding Dojo’ - programming club at DKTES.
  • Organized ‘Code-A-Thon 2k19’ a national level and ‘Coding Dojo 2k19’ an institute level programming event respectively at DKTES
  • Conducted workshops on Python programming which helped non-technical students to crack TCS assessment.
  • Created an AI repository on GitHub which helps students to learn the basics of Python and ML.
    Github Link: https://github.com/dktes/moodle.ai

Social Media

Contact