Avinash Kanumuru, Developer in Bengaluru, Karnataka, India
Avinash is available for hire
Hire Avinash

Avinash Kanumuru

Verified Expert  in Engineering

Bio

Avinash has over 13 years of experience in data science and artificial intelligence, specializing in driving the development of scalable data and ML services within the dynamic landscape of fintechs. Driven by an insatiable curiosity and a relentless pursuit of excellence, Avinash is deeply passionate about leveraging emerging technologies to deliver automated solutions that address pressing business needs while continuously expanding his skill set to remain at the forefront of innovation.

Portfolio

Standard Chartered
Python 3, Data Science, Machine Learning, Artificial Intelligence (AI)
HSBC
Random Forests, Python 3, Machine Learning
Schlumberger
Data Analysis

Experience

  • SQL - 13 years
  • Data Science - 10 years
  • Pandas - 8 years
  • Python - 8 years
  • Scikit-learn - 8 years
  • Artificial Intelligence (AI) - 8 years
  • Machine Learning - 8 years
  • Deep Learning - 5 years

Availability

Part-time

Preferred Environment

Git, Linux, Python 3, Visual Studio Code (VS Code), MacOS

The most amazing...

...project I've developed was a machine learning algorithm to extract transaction tables from PDF bank statements and categorize them for credit underwriting.

Work Experience

Manager | Advanced Analytics AI/ML

2021 - 2022
Standard Chartered
  • Designed and built an in-house modular solution in Python for AutoML covering all aspects of MLOps, from data preparation to model explainability and packaged deployment.
  • Built the algorithmic decision framework for credit cards and unsecured lending using an in-house AutoML solution.
  • Led data scientists in building predictive models like engagement score, segmentation, and churn models.
Technologies: Python 3, Data Science, Machine Learning, Artificial Intelligence (AI)

Senior Manager | Decision Science

2014 - 2021
HSBC
  • Developed an ensemble classifier to categorize transactions into various cash flow heads and assess customers' repayment capability for lending purposes.
  • Developed the capability to extract transactions from PDF bank statements.
  • Contributed to a customer micro-segmentation project using the HDBSCAN technique to target niche segments for personalized messaging.
Technologies: Random Forests, Python 3, Machine Learning

Research Analyst

2013 - 2014
Schlumberger
  • Provided consultants with operational and financial analysis for cost-cutting measures in oil and gas upstream and participated in consulting engagements with big clients.
  • Benchmarked SBC performance in social media with peer and competitor firms and recommending action plans.
  • Suggested and implemented process improvements that automated processes.
Technologies: Data Analysis

Senior Software Engineer

2009 - 2012
Wipro Technologies
  • Created an eCommerce project for medical products category for US clients including Microsoft's piracy control team and Cardinal Health.
  • Produced business intelligence reports for fraudulent usage of activation keys using a mix of threshold and policy-based rules as part of a DWH-BI project.
  • Developed dashboards for monitoring databases for teams and automated generating resource billing status reports for senior management.
Technologies: Data Warehouse Design, Business Intelligence (BI), SQL

Experience

Detection of ID Card and Extract Details

Developed a machine learning model to detect ID cards and then identify types of ID cards using a trained model. Further extract details from the ID cards to autofill product application forms.

Achieved an accuracy of 87%, including extracting correct details from the ID cards. The model is trained on a set of four types of ID cards with various resolutions and orientations.

Complaint Categorization Using Topic Modeling

Categorized complaints into various topics (concern areas) using NLP algorithms (LDA and LSTM). Further performed sentiment analysis to find the customer emotion and rank each complaint accordingly to help business with customer satisfaction KPI metrics.

Customer Segmentation with Density-based Techniques

Segment customers to find behavioral micro-segments based on demographic and transaction activities and communicate relevant messages or offers, using machine learning algorithms like k-means, k-NN, and HDBSCAN techniques.

Education

2012 - 2014

MBA in European Business

ESCP Business School - Paris, France

2012 - 2012

Post Graduate Diploma in Management (PGDM) in International Management

Management Development Institute (MDI) - Gurugram, India

2004 - 2008

Bachelor's Degree in Electrical and Electronics

Osmania University - Hyderabad, India

Certifications

MAY 2021 - PRESENT

Tableau: Hands-on Tableau Training for Data Science

Udemy

JANUARY 2021 - PRESENT

Spark for Machine Learning and AI

LinkedIn

DECEMBER 2020 - PRESENT

PyTorch Essential Training: Deep Learning

LinkedIn

OCTOBER 2019 - PRESENT

Agile Foundations

Project Management Institute (PMI)

FEBRUARY 2016 - PRESENT

Machine Learning by Andrew Ng

Coursera - Stanford Online

Skills

Libraries/APIs

Pandas, NumPy, Scikit-learn, XGBoost, OpenCV, PySpark, REST APIs, Spark ML, PyTorch

Tools

GitHub, Seaborn, Git, BigQuery, Tableau

Languages

Python 3, SQL, Python

Platforms

Jupyter Notebook, Amazon Web Services (AWS), Visual Studio Code (VS Code), Google Cloud Platform (GCP), Linux

Frameworks

Spark, Apache Spark

Paradigms

Business Intelligence (BI), ETL

Storage

MySQL

Other

Data Science, Data Analysis, Random Forests, Logistic Regression, Machine Learning, Artificial Intelligence (AI), Business Strategy, Data Warehouse Design, Natural Language Processing (NLP), Data Analytics, Statistical Modeling, Generative Pre-trained Transformers (GPT), Business Management, Agile Practices, Deep Learning

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring