Skip to content

About Manikanth Reddy

Hello there,

Thanks for taking time to know more about me. My name is Manikanth Reddy (you can call me Mani). You can contact me through the information provided in this page.

Disclaimer:

This is my personal website and the views expressed here are mine alone and do not necessarily reflect the views of my employer, unless and until I mention it in the corresponding page. Please go through the full disclaimer here.

Summary

Currently I am working as a Senior Machine Learning Engineer (Tech Lead) at Bloomreach. I am working on developing and deploying data science products for Bloomreach customers at scale. You can use some of the products I worked on, on websites like Puma, Albertsons.

Before I joined Bloomreach, I worked as an Data Scientist (Machine Learning) at American Express. I was a part of Machine Learning and Data Science (MLDS) Research team based out of Bangalore, India. As part of the team, I worked on solving a broad range of problems related to Credit and Fraud Risk.

In my free time, I work on Kaggle, go on trips (Bike trips are ❤️) or read manga (can talk days about One Piece, Haikyuu and Jujustu Kaisen, if you are into it). I also play Table Tennis, Cricket, Badminton and Volley Ball. I also developed building Video Games (mainly in Unity) as a hobby.

Interests

My current ML research interests include:

  • Recommendation systems
  • Multi-modal representation learning
  • Natural language processing and language models
  • Self-supervised and semi-supervised Learning

I am also interested in:

  • Even through I work as an ML Engineer, I am very much interested in desiging and developing non-ML systems as well. I plan to work at an early-stage startup or start my own company in the future.

  • For the early-stage startups, I would only prefer the companies which:

  • Are working on a mission which I believe in.

  • Have a great team and culture.

Bloomreach

As an ML Tech Lead at Bloomreach, I am responsible for:

  • Leading the design and deployment of recommendation models for our customers.
  • Participate in, and leading ML/AI design reviews with peers and stakeholders to decide amongst available technologies.
  • Mentoring and guiding junior ML & software engineers - reviewing their models & code and to provide feedback to ensure best practices.
  • Working cross-functionally with product managers, engineers, and program managers to develop and deploy AI/ML solutions.
  • Developing Complete the Look product which allows our customers to suggest automated looks for products to their users. This is currently manually done by fashion experts and doesn't scale well.
  • Developing Shop The Look (or Visual Search) product which allows our customers (merchants like Puma, Albertsons etc.) to allows their shoppers to search using Images on their websites. It will be similar to Google Lens or Amazon's Shop The Look, but more towards Apparel, Fashion and Home Decor.
  • Improving the existing Similar Product Recommendations (called More Like This or MLT) using Deep Learning. We are using Transformers based BERT variant model. This product was launched on March 23, 2022 and is available in production to be used by our customers.
  • Previous I have worked on developing Attribute Extraction (AE) model for German and French languages. Our search uses the tags generated by the AE model for providing semantic search capabilities. Click to view about the project on our company blog.

American Express

My last project at American Express was on Bank Statement Summarization using Natural Language Processing.

  • I helped reduce the response time of the ML Project by 3 fold using Code Optimization and Caching techniques which led to satisfying SLA time requirements for the first version of the model.
  • I am currently responsible on building models for International markets and Unified model for all languages.
  • I have been working on this project for last 3 months and due to project restrictions I can't discolse any more details.
  • I am sure that this is one of the best projects I have worked on till now. I got the chance to enchance my technical and business skills:
  • Technical Skills - Being my first big NLP project, I got good pratical experience from basic embeddings to using latest SOTA techniques in Natural Language Processing. Understood how to use productionize models using Docker. Working on the production code base with colleagues.
  • Business skills - First time working on a project which involved more than 50 colleagues working on the same project. Got good understanding of how to present to various stakeholders with varying bussiness and technical skills (From VPs to Analysis) as well as large audiences.

As part of the Machine Learning and Data Science team, I have also worked on:

  • Creating a loss function to solve multiple business objectives (Capture rate and Dollar Savings of Credit and Fraud Models) at the same time.
  • Upgrading RNN Suite - A package used across American Express modelling teams to build Time Series based Machine Learning models for various usecases.
  • Revenue forecasting of companies using XGBoost on sparse data.

  • Some other research projects I have worked on include Neural ODE (Ordinary Differential Equations), Adverserial Attacks on Machine Learning Models for Fraud Detection (lasted only for few days), Knowledge Distillation, Attention based Neural Networks.

  • Also as part of the Machine Learning and Data Science team I have presented on the 3 research learning sessions to the Bangalore Decision Science teams:

  • PyTorch Training (From Basic to Advanced)
  • Top Research Topics from MLDS 2020 Bangalore Conference
  • Knowledge Distillation

I have received SVP Star Acheivement Award for Q4 2019 (Decision Science & Data Strategy Team) and I believe that I am one of the youngest to the receive this award...

IIT Kharagpur

I recevied my Bachelors Degree from Indian Institute of Technology, Kharagpur. My major is on Metallurgical and Materials Engineering. I was ranked 3rd among Bachelor Degree students in my department.

  • I have participated in various Hackations:
  • I got 2nd place in Data Science Hackathon conducted by HSBC (5 man team) and was invited to HSBC Data & Analtyics Symposium 2019 held at Kolkata.
  • I was placed in Top 10 teams of AnalyzeThis 2018 competition held by American Express and received a PPI (Pre-Placement Interview Offer) for the same. I was converted this offer into full time and Joined American Express in July 2019 as a part of Machine Learning and Data Science Team. (2 man team).
  • I participated in Code.Fun.Do 2015 competition held by Microsoft and built an app within 2 days in C# (3 man team).
  • I learnt Programming and Data Science skills from Coursework and MOOCs.
  • Some Relevant Coursework Include: Design and Analysis of Algorithms (Theory and Lab), Programming Data Structures, Artificial Intellience, Soft Computing Tools (Machine Learning, Fuzzy Logic and Evolutionary Algorithms), Probability and Stochastic Processes, Computer Applications in Metallurgical Processes.
  • Some MOOCs I have took include: Machine Learning, Applied Data Science in Python, Deep Learning, TensorFlow for Deep Learning Research, Deep Learning for Natural Language Processing.
  • I was the Web Development Lead for COMPOSIT. I lead a team of 5 people to build the Website for the 25th edition of the conference (2017-18).
  • As a Dakshana Alumni (DAAN Alum, an NGO), I have volunteered to organize Grand Dakshana Selection Test for 3 years. I was also the college representative for 2016-17 sesion.

Jawahar Navodaya Vidyalaya

I completed by 12th Standard from Jawahar Navodaya Vidyalaya, Gachibowli as a Dakshana Scholar. I become a Dakshana Scholar after clearing the Grand Daskhana Selection Test 2013. Dakshana is an NGO registered in India.