Highlights:

Skills:Machine learning

Senior MLOps Engineer 08/2024 - current
Accenture Poland, warsaw, Warsaw Poland
Here I was responsible for integrating model monitoring in the existing pipeline for an Accenture client
Client: Mondelez International ? Developed re-usable modularised scripts for data validation, data drift and model drift monitoring on Databricks utilising custom code and whylogs framework. ? Integrated the monitoring process with the existing MLOps pipeline workflow deployed on Databricks along with automated email notification in case of drift detection. ? Created a guided documentation to extend similar monitoring process to other internal projects at Mondelez--
MLOps Engineer 11/2022 - 08/2024
Accenture Baltics, Riga, Riga Latvia
For Accenture Baltics, I worked for two clients: TE Connectivity, USA and Piraeus Bank, Greece. Both roles are in the area of MLOps and model productionisation on AWS, Azure and Databricks cloud
Client: TE Connectivity ? Built an AWS cloud machine learning platform that enables developers to create, train and deploy machine learning models into production followed by CI/CD pipeline using MLOps principles. Thereby reducing time to productionize to 2 weeks. ? Build ML system architecture for deploying models on Amazon Web Services (AWS) followed by detailed presentation to internal stakeholders. ? Designed and implemented automated workflows for data preprocessing, model training, and deployment using AWS Lambda, CI/CD pipelines, CloudFormation, Step Functions CodePipeline, and Docker, reducing deployment & re-training time by 50%. ? Collaborate with Data Scientists, Data Engineers, Leaders, and other IT teams to build products solving strategic problems for multiple Business Units (BUs) like Marketing, Sales, Supply Chain etc. ? Worked with data scientists to improve the efficiency of the machine learning models and streamline the model deployment process, resulting in a 25% increase in model throughput. ? Configured SNS notifications for successful, failed model deployment and sending appropriate model output status to downstream analysts, resulting in a 20% reduction in downtime and on-time notification. Client: Piraeus Bank – Public Sector Bank ? Built an end-to-end DataOps platform to deploy data engineering workloads on Databricks using bash scripting & Azure DevOps pipelines. This includes automatic notebook deployment, cluster creation and job deployment using Databricks CLI and REST API. ? Deployed ML and deep learning models in the area of predictive modelling, computer vision and recommendation system on Azure cloud following CI/CD pipeline. ? Deployed a deep learning GPU accelerated batch transform image processing model on AWS followed by CI/CD pipelines using AWS CodePipelines, CloudFormation, SageMaker, Docker and AWS Lambda. ? Built Azure Databricks MLOps platform for deploying MLflow pipelines across 3-environments utilising Azure Bicep and Azure DevOps pipelines. ? Developed a roadmap for end-to-end MLOps serverless system design for creating sentiment analysis using OpenAI GPT-4 API. ? Provide guidance to fellow data scientists and ML engineers to follow best practices for code refactoring and repository management throughout the whole software development lifecycle.--
Machine Learning Engineer 07/2021 - 07/2022
Blackstraw AI, Chennai, Tamil Nadu India
Industry: Information Technology
In this role, I am responsible machine learning model development and deployment as well. I am also responsible for managing and creating CI/CD pipeline on Azure cloud.
? Developed predictive interest model to predict interest of medical clinicians in the job and predictive offer model to predict likelihood of getting a job offer, followed by CI/CD MLOPS pipeline on Azure DevOps. Thereby, reduced manual screening time from 2 hours to 6 minutes. ? Recency, Frequency and Monetary (RFM) analysis was done for client clustering followed by statistical Lifetime Value (LTV) modelling to aggregate most profitable hospitals for AMN Healthcare, USA. ? A generic reusable RFM modelling product is developed which would save development cost of such product by 80% for other potential clients in future. ? Developed a Priority Scoring methodology that prioritise important & profitable clients/hospitals using XGBoost algorithm, followed by modular and production ready code. This scoring helps to retain and prioritise hospitals which are responsible for 70% of the revenue generated. ? Delivered internal sessions to fellow data scientists on continuous integration MLOps pipeline to make data scientists self-reliant towards CI/CD practices. Thus reduced the model development and testing time in the production environment by 50% followed by reduction in technical dependency on the DevOps team. ? Refactored existing machine learning legacy code to production ready, reusable, modular and scalable code using python OOP standards to maintain code quality & reusability. Such code reusability cut down cost & technical debt by 50%. ? Developed and delivered PoC presentation to McDermott International, USA at Executive level for the development of end-to-end machine learning solution framework for predicting time of arrival of container ships in the port for offshore construction of oil and gas refineries. ? Developed ETL script for extracting 100 GBs of data from Snowflake followed by data transformation and loading into Snowflake for power bi dashboard development.--
Data Analytics Consultant 09/2020 - 08/2021
QUALITY COUNCIL OF INDIA, MINISTRY OF COMMERCE, New Delhi, Delhi India
Industry: Government Administration
This is a consulting role where I am responsible for implementation of tourism intelligence system for Ministry of Tourism.
? Developed a regional tourism monitoring PoC for the Ministry of Tourism for post COVID-19 planning & supplementing tourism activities in India. ? Time Series Forecasting & Predictive Modelling for tourist footfall determination. ? Text Modelling, Sentiment Analysis & Named Entity Recognition (NER) is performed for improving sightseeing experience and simplifying tourist feedback & complaint system. ? Extracted historical tourism data from various government stakeholders and loaded into central Azure Blob Storage. ? Collaborated & coordinated meetings with Director General of Tourism and technical consulting partner for strategic onboarding for the scaling tourism intelligence PoCs to national level. ? Performed comprehensive compliance analysis on Google, Facebook, Amazon, and Twitter as a part of antitrust cases in India for following non-competitive practices on the recommendation of the Ministry of Commerce & Industry. ? Scrapped 30,000 products dataset from E-commerce giants (~85% market share) in India for the Ministry of Consumer Affairs and performed regulatory & compliance analysis under Consumer Protection Act (E-commerce Rules) 2020. ? Liaised with diplomatic stakeholders to deliver complex analysis in a clear, concise manner to be actioned in a way suited to both technical and non-technical audience. ? Contributed to the development of RFPs & winning new contracts followed by relevant data gathering, technical framework design and data verification to address business issues.--

Aligarh Muslim University 08/2015 - 06/2019
Aligarh, Uttar Pradesh, India
Degree: Bachelor's Degree
Major:Mechanical Engineering
4 year undergraduate full-time program in mechanical engineering and technology

Login to view resume: MohdDanish_Resume -