Avatar

Software Developer & Sketcher

Namaste! I'm Avinash, software engineer @ Innovaccer and graduate in Information Technology from IIIT Allahabad. I am interested in Software Development, Web Development, Deep Learning, and GenAI solutions.

Get in Touch

Technical Skills

I'm working in the domain of Software Development, AI, and web development. I'm looking forward to finding and optimizing the solutions to problems that can be tackled.

Python Logo

Python

Django Logo

django + django REST framework

Flask Logo

Flask

c++ Logo

C++

Java Logo

Java

git logo

Git

MySQL Logo

MySQL

PostgreSQL Logo

PostgreSQL

Elasticsearch Logo

Elasticsearch

Terminal Image

Command Line

Docker Logo

Docker

Apache Spark Logo

Apache Spark

Experience

Some of the professional positions I've held.

Innovaccer
Jan 2023 - Present

SDE II @Innovaccer Middle East, Abu Dhabi

  • POC on various speech-to-text options like Google Speech-to-Text AI, selected whisperX for its multilingual and medical terminology capabilities. Integrated whisperX into Sara, a doctor-patient conversation recorder, reducing documentation time by 50% and improving accuracy, for generating SOAP notes using the GPT model.
  • Developing Provider Relationship Management (PRM) application with FastAPI for backend services, and Flutter for frontend.
  • Created MVPs for Multi-Tenant User Management, Market Intelligence, and Network Optimizer using Django Rest Framework and React/Next.js.
  • Automated the package upgrade process using Mend Renovate, leading to a remarkable 80% reduction in CI pipeline failures during vulnerability scans.
  • Complete ownership of developing package management system to regulate Python, JS, and Java packages through a private artifactory, resulting in 99% improved security.
Innovaccer
Oct 2022 - Dec 2022

SDE I @Innovaccer Middle East, Abu Dhabi

  • Latency Optimisation by 50% through code optimization and migrating patient-incentives app from Django to FastAPI framework.
Innovaccer
July 2021 - Sept 2022

Software Development Engineer ‑ I @Innovaccer, Noida

  • Played a key role in developing the AIMS web app, working in partnership with Roche and utilizing FHIR resources.
  • Collaborated closely with the QA team to enable client-independent data creation, offered support for business logic, and assumed responsibility for implementing essential enhancements from a development standpoint.
  • Optimised latency by 20% through the improvement of database queries and the implementation of architectural enhancements.
Explore ML
July 2019 - May 2020

Facilitator @Google's Explore ML

Explore ML is a Google-sponsored program for the university to get started with machine learning.

  • Facilitated 3 workshops on ML at IIIT-A

Projects

Some personal and college projects

View GitHub

Project
Description
Spark MovieLens
  • A System for Movie Recommendation using a content-based, collaborative approach on the static data set.
  • Extended to real-time as a web portal using Flask, Apache Spark, and MLlib.
  • Technologies Used: Python, Apache Spark (distributes computing), Flask.
  • Real-Time Smart City Map on (Kumbh) Temporal Data
  • An Android app has features of maps that google maps can’t reach such as a temporal city like Kumbh.
  • KML(Keyhole Markup Language) is used to display geographic data collected through GPS.
  • It is designed for both the admin and user side where the government (admin) can add/update any protocol and the user can get instant notifications.
  • Technologies Used: Android Studio, Firebase Realtime Database, KML.
  • Data Mining Approaches in Healthcare Operations to Improve Decisions and Activity Workflow
  • The principal objective of this work is to solve two operational issues in healthcare management: predicting operational activities delay, and recognizing key work process drivers.
  • Technologies Used: Python, data mining.
  • Image Super-Resolution using GAN
  • A Generative Adversarial Network has two main Neural Networks, Generator, and Discriminator.
  • The Generator takes a low-resolution image as input and tries to generate a high-resolution image for that input. The Discriminator learns from the actual training data and also the generated results of the Generator and tries to distinguish real and generated images
  • As the training goes the generator gets better and better at generating realistic outputs while the discriminator finds it harder and harder to distinguish real and fake images.
  • Technologies Used: Python3, GAN
  • Resume

    I'd love to hear from you, email me at avinashyadav06.ay@gmail.com

    Wanna know more?

    RESUME