Achyudh Ram

Hi! I am Achyudh Ram, a visiting scholar at the Software Architecture Group, University of Waterloo, working on the empirical modeling of sentiments in pull request discussions. A few months back, I was at the Software Engineering Research Group, Delft University of Technology (TU Delft). I also had briefly interned as a Software Developer at Intuit Inc.

I am interested in building intelligent software development automation systems that can perform high-level cognitive tasks, by bringing together diverse fields like software analytics, machine learning, information retrieval and social networks.

I write about some of the interesting aspects of my projects in my blog. If you are after a comprehensive list of my work, take a look at my curriculum vitae or GitHub profile.


Investigating Type Declaration Mismatches in Python

In Proceedings of MaLTeSQuE 2018 (Workshop on Machine Learning Techniques for Software Quality Evaluation)

Blog posts

Why study the reviewability of code changes?

I set out to find what contributes to the “reviewability” of pull requests submitted to OSS projects.

Propr: Assess pull request reviewability

Why did I develop Propr, and how can it help developers improve their pull requests?

Bringing Network Analysis to Software Engineering

I was excited to work on a project with the Networks Group at BITS Pilani, and had no idea what would ensue.

Recent projects


Empirical modeling of developer sentiments by learning from user discussions on GitHub with deep neural networks.


An automated tool to identify type inconsistencies between source code and method docstrings in Python across popular Python libraries.


An automated evaluation framework to assess the reviewability of pull requests, that is modularized using container-based virtualization.


A suite of developer feedback and report generation tools for the identification of factors associated with the reviewability of pull requests.


A feed-forward Neural Network library using computational gate approach supporting multiple optimizers, common activation and loss functions


An analysis framework to decode the social structure of developer mailing lists, IRC channels and Slack teams.

FCM Feature Selection

A novel feature selection technique using cosine similarity scores on the semantic centroids calculated from the normalized term-term correlation factors based on Fuzzy C-Means clustering.

ASL Translation

An ensemble classifier for translation of American Sign Language gestures with hard negative mining and non-maximal suppression for localization.

Legislative Trends using NLP

Identification of latent structures within Indian parliamentary debates using natural language processing to discover seasonal trends in the debates of the upper and lower houses.

Get in touch!

Send me an email or tweet to me. You can also check out my profile on LinkedIn and GitHub.