Manoj Kumar Jha, Ph.D.

Æ +1-470-***-**** [ ad18ie@r.postjobfree.com LinkedIn Atlanta, Georgia, USA Summary

A collaborative professional with substantial experience in designing, developing and executing solutions for Big data using public and hybrid clouds for financial and scientific domains. Expertise in using the right tools and creating an intuitive architecture that helps organizations with ingesting and analyzing Petabytes of data and hence increased customer satisfaction.

Education

University of Delhi Delhi, India

Ph.D (Experimental Particle Physics) March 2007

Master of Science (Physics) March 2000

Work Experience

Equifax Alpharetta, Georgia, United States

Senior Software Engineer in Big Data Nov 2020 – Present, Full-time

• Took a leadership role in defining the lifecycle of the Machine Learning (ML) project. This includes defining the goals of the task (Scope), getting right data for the training model, iteratively refine the model using data centric approach, deployment and finally montior the performance of the model. For deployment, we used common and well tested deployment patterns like canary and blue/green approach. Used Dashboard for monitoring the performance of the model. Following type of metrics are collected: software metric (memory, compute, latency, throughput), input metrics (like number of missing values), output metrics ( returning null value).

• Design and build high frequency and availability, low latency address standardization service deployed in Google cloud. Used Istio as service mesh for apps to be resilient, secure, and observable as well as reduce the risk of shipping new features. Tool Kubernetes is being used for container orchestration.

• Build data processing pipelines for seeding records using Apache Beam and Dataflow runner.

• Mentor junior engineers in software development, technology and processes.

• Troubleshoot complex issues discovered in-house as well as in customer environments. Used metrics collected (like average number of clients, CPU and memory consumption, HTTP response stats, ...) from monitoring dashboard

(like Grafana, Stackdriver, Kubernetes cluster page) for debugging the business units issues.

• Participated in code reviews and release management activities. Equifax Alpharetta, Georgia, United States

Software Engineer in Big Data August 2016 - Nov 2020, Full-time

• Used Cloudera Data Platform for building next generation core exchanges for Equifax. Incoming Data was extracted, transformed using Spark processing engine. Data was stored in Hadoop Distributed File System. Apache Hive was being used for data driven decisions.

• Splitted the monolithic batch fullfilment service into smaller number of microservices. Visualize the use case which generates and responds to a continuous stream of events. Reframing the business in terms of a continuous stream of events offers huge benefits like freshen insights, a single version of truth, faster reaction time and simplified architecture. Deployment of these microservices based on this pattern results in reduction of 50% of response time. Microservices were written using Spring boot framework and deployed in service mesh offered by Istio.

• Implemented fine grained authorization policies across several business domains using Open Policy Agent.

• Work seamlessly with Agile development team.

Purdue University West Lafayette, Indiana, United States System Analyst for CMS Experiment Oct 2012 - July 2016, Full-time

• Used Hadoop Distributed Filesystem as storage and analysis of data. HDFS was able to store and analyze several Petabytes of data.

• Keep the data separated and secured across national boundaries through multiple data centers using public key X509 infrastructure.

• Work with data and analytics experts to strive for greater functionality in the data systems.

• Implemented and configured HDFS for performance at scale. INFN-CNAF and CERN Geneva, Switzerland

Application Developer and Distributed Computing Project Support September 2009 - September 2012, Full-time

• Build analytics tools that utilize the data pipeline to provide actionable insights into customer acquisition, operational efficiency and other key business performance metrics

• Developed the experiment Dashboard that offers a solution to monitor and manage jobs running across several computing sites. This interface is being used for monitoring work flows and health of the computing sites [1].

• Experience in writing experiment specific plugins to allow seamless integration of the experiment software to be used for distributed computing resources. Implemented strategies for splitting large computing jobs to allow faster parallel processing [4].

• Developed the automated calibration framework for the data coming from the LHC detector. Transformed

(calibrated) data sent back to CERN for use in analysis and production jobs [5]. INFN and Fermilab Bologna (Italy) and Batavia, Illonois, USA Grid Middleware Developer September 2007 - August 2009, Full-time

• Proposed and implemented a new data transfer model that uses Storage Resource Manager (SRM) as local caches for remote Monte Carlo production sites, interfaces them with experiment data catalog (SAM) and finally realizes the file movement exploiting the features provided by the data catalog transfer layer [6].

• Demonstrated quantitative and analytical skills in mining extremely rare physics process signals from large multi-component noise using statistical modeling techniques. Skills

Cloud services: Google cloud platform (Dataflow, Dataproc, Pub/Sub, Datastore, CloudSQL, K8s, Stackdriver alerting and monitoring.)

Security and encryption tools: Vault, Public key X509 infrastructure, Kerberose, SSL. Computer programming languages: Java, Go, C++, Python, Shell script, Fortran. Database: SQL and NoSQL.

Debugging tool: Valgrind, strace, pdb and gdb.

Distributed data processing: Hadoop ecosystem tools, Spark, File system: HDFS and GPFS.

Version control systems: Git, SVN and CVS.

Builds and dependency management systems: maven, gradle and bazel. Configuration management: Puppet.

CI & CD systems : Jenkins.

Metric systems: Grafana, Prometheus, Stackdriver

Container Orchestration:Docker, Kubernetes (K8s).

Service Mesh: Istio.

Visualization tools: Matplotlib, ROOT.

Machine Learning:Classification, Regression

Platforms: Linux, Windows and OS X.

References

[1] “Experiment Dashboard Task Monitor for managing ATLAS user analysis on the GRID”, J. Phys. Conf. Ser. 513 032083, 2014.

[2] “ATLAS computing activities and developments in the Italian Grid cloud”, J. Phys.: Conf. Ser. 396 042052, 2012.

[3] “Multicore in Production: Advantages and Limits of the Multi-process Approach”, J. Phys. Conf. Ser. 368 012018, 2012.

[4] “Reinforcing user data analysis with Ganga in the LHC era: scalability, monitoring and user support”, J. Phys. Conf. Ser. 331 072011, 2011.

[5] ATLAS Muon Calibration Frameowrk, J. Phys. Conf. Ser. 331 072007, 2011.

[6] “A new CDF model for data movement based on SRM”, J. Phys. Conf. Series 219:062052, 2010. Last updated: December 23, 2023

Contact this candidate