Senior Product Engineer FinTech Backend Systems

Mithu Ganesan

Phoenix, USA *****

602-***-**** - ************@*****.*** - WWW: https://www.linkedin.com/in/mithu-ganesan-b8274433 Work Permit

• Visa Status: US Citizen

Summary

Dynamic Product Engineer with over 15 years of experience in developing high- performance, scalable backend platforms in FinTech, payment processing, and enterprise retail systems. Expertise in C++, Go, and Java, with a specialization in low- latency distributed systems and RESTful API design. Proven ability to manage features end-to-end, modernize legacy platforms, and support millions of transactions at peak scale while ensuring reliability and security. Committed to leveraging collaborative approaches to enhance product functionality and user experience, consistently delivering high-quality engineering solutions through innovative problem-solving. Technical Skills

Programming languages: C, C++, Perl,

Shell scripting, Python, Java,

JavaScript, Go, PHP

• • Network routers and file servers

Experienced with Fedora, RHEL, and

UbuntuMate

• Tools: Gitkraken, ActivePerl IDE,

Eclipse IDE, Selenium, GDB,

Valgrind, cvs, git, Bugzilla,visual

studio code,cmake, clang-tidy, redis,

JIRA, Confluence, Slack, Bitbucket,

Kafka lenses, Splunk, oneOPS, APIC,

SOAP UI, POSTMAN, IntelliJ,Tomcat,

Maven, Automation,

kibana,Cassandra,informix, DB2

•

• Analysis of client requirements • Prototype development and evaluation

• Root cause analysis • Proficient in finite element analysis

• Process efficiency improvement

Work history

07/2020 to 01/2026 SENIOR PRODUCT ENGINEER

EPX – TEMPE, AZ

Created a new application EPI, which converts json requests from web clients into an XML request which is then sent for authentication and processing. Receives XML request and converts it back to json for the client. Mainly used by web based

•

RESTFUL clients. EPI is also RESTFUL where the url has options to supply parameters. Also pulls bulk data from oracle DB and omits null values using XML Element and XML forest before convert to json and sending for auth. Endpoints are defined with specific numerals in DB, based on the endpoints specific functionality is performed.

Created a helper for EPI, purely in GO. This actually helps EPI with the conversions of the requests, from XML to Json and vice versa.

•

• EPI – Electronic Payment Interface

• EPI Adapter

05/2020 to 02/2023 SENIOR SOFTWARE DEVELOPER

WALMART – BENTONVILLE, AR

Bulk Transactions are handled using web(using REST) or as batch process. Batch process indicates, creating a batch file with all the transaction happened during specific window or interval of time and processing the file as one transaction. Considered efficient way to reduce DB traffic, network latency. All the transactions are processed and published as Kafka Topics and Updated back to Cassandra which is then consumed by other applications.

•

Worked on Walmarts retail ecommerce wing, and supported the e-commerce application team, where all the smart store data is transferred through rest services or batch files to e-commerce site. Supported during thanksgiving and other holiday events and handled millions of traffic based on the observability tools like splunk, oneops, graffana, azure server storage and kubernetes containers.

•

All the data is transferred to Cassandra or published in Kafka based on e- commerce customer needs. From where Walmarts internal e-commerce customers and scan and go systems pull the product data (price, inventory, basic item info, cross reference items, offer items, backroom data, location data). Cassandra clouds are monitored for RPM(request per minute) and observed for any errors or spike in traffic or unusual response codes. Grafana is also used to observe Cassandra and oneops cloud(where the application is running through tomcat). Kafka traffic and errors are observed through lenses platform. Also log into the console of the IPs available in oneops cloud and observe the transactions (audits logs) for unusual load and enable more clouds based on the need / traffic or intimate the consumer about the overload.

•

Worked on agile environment where we had daily stand ups with all team members including offshore. While going over stories in each team members lane in JIRA, and tracked the efforts for every sprint. Allocate the projects and efforts for each team member and rotate the work so everyone has hands on, on all the work. Present the teams effort to top leadership every month so efforts are tracked, and projects are acknowledged.

•

Created few helper applications for the ecommerce wing, which includes batch elimination for certain countries. In this all the above-mentioned process of creating a batch file to process all the transactions is removed and everything is processed through WEB via REST API. initially there was a threshold after which

•

(30000) all of them are processed as batch(file). Now with batch elimination the threshold is removed, and the traffic is handled by load balancer.

• Batch Process

• Observability and Monitoring

• Batch Elimination

05/2017 to 03/2020 SENIOR SOFTWARE DEVELOPER

EPX – TEMPE, AZ

Created an application which loads data from database tables and columns into redis (which is an In memory map), Based on time interval, older data is wiped and new data gets added in the form of hashmap(key – value pairs). Stored procedure is executed in database on specific intervals, based on the status of each transaction CRUD(create, read, update, delete) operations are carried out in redis. Applications using redis instead of database for backed up data are much faster and efficient. This is can avoid network latency and makes application 90 times faster for pulling huge data. Used OCCI's data buffers to pull huge data from database. All Merchant information, card information for different networks(AMEX, VISA, MasterCard, Discover), transaction tables etc., all are moved to redis and specific type of service is enabled based on the needs.

•

Created an application using core Java, which loads files from different networks to be processed for settlement in batches. Also Created an application which creates response files which has to be sent back to networks(AMEX, DISCOVER

,VISA,MASTERCARD)

•

EPX(electronic payment exchange) is a transaction processor that works as middle layer between various networks and acquirers(banks) The Sales Data files provides detailed information regarding the submission of Sales Data to respective networks' Card Transactions. Used ZeroMQ sockets for queuing and client-server programming Used OCCI's bulk retrieve/update model to retrieve and update 1000 to 10000 rows at a time (in milliseconds) and created a formatted file as per the networks'(Discover, Visa, Amex) specifications. Used Multi threading, smart pointers, and few C++14 features for efficiency.

•

Developed a caching tool, which uses sqlite3 to store tables in hard disk. Whenever there is more frequency for particular rows of the table being used, moved them in memory map structure and used a timer class to update the map every minute to clear old data(not used in last 60 seconds). This way maps size is well maintained. This tool reduced network latency for many applications for accessing database tables. Used ZeroMQ sockets for queuing and clientserver programming.

•

• RedisSync – EPX

• File Loader And Response File Creator Using Java – EPX

• Creating Sales Data Files for Networks(Visa, Amex, Discover, MasterCard) – EPX

• DBCache – Caching Database tables locally to reduce network latency

• Other Works

11/2015 to 05/2017 SOFTWARE DEVELOPER

CROSSOVER SOLUTIONS – REMOTE

• Created a high performance RESTful web service.

National stolen vehicles database contains information about stolen vehicles on a cloud hosted Mongo DB database using tables and views like Vehicle, Fields: Vehicle Registration, Vehicle Make, Vehicle Model, and Vehicle Owner.

•

Created a RESTful web service that would allow searching for vehicles in this database. The search can be made in all the data fields of the tables created.

•

This service will also log all the searches made into another table named “log” in the same database.

•

Crossover is a freelancing company which offers remote jobs and works with remote teams.

•

• National Stolen Vehicle Database Management

03/2013 to 06/2014 ASSOCIATE SOFTWARE ENGINEER

IBM LABS – BANGALORE, INDIA

Team develops purpose-optimized Linux operating environments for use as embedded OS for IBM products and I have worked on functional integration verification, wrapper layer development, network analysis like InfiniBand and TAHI, stress testing, bash scripts, root-cause analysis and maintenance.

•

IBM software Labs deals with maintaining the IBM servers, Operating systems running on the servers, open source Linux projects, RHEL, fedora and other Linux flavors

•

• IBM Embedded Linux OS functional integration

07/2011 to 03/2013 SOFTWARE DEVELOPER

CISCO SYSTEMS – BANGALORE, INDIA

Ported SNMP modules from old router operating systems (QNX- Neutrino) to new generation of routers which are completely based on Linux (fedora) Watchdog development and monitoring - CISCO Systems

•

Watchdog System Monitoring (wdysmon) is a process which monitors all the process that are registered with it, for every 60 seconds interval and generates a process crash when any of the processes fails to give signal to it. I have developed a feature to write the state of the crashed process to a file with analysis about the current state of the router at which the crash happened. This file can later be used to debug the process which crashed and to know the state of the system during the crash.

•

• SNMP OS MIBS porting to next generation (NG) routers

• Watchdog Development and Monitoring

01/2011 to 07/2011 MEMBER TECHNICAL STAFF INTERN NETAPPS – BANGALORE, INDIA

E-Passport is a RFID based passenger authentication system. Every passport is embedded with a RFID tag having a unique EPC (Electronic Product Code). RFID reader is placed at the entrance and exit of the airport. When a passenger enters the airport, reader scans the RFID tag embedded in the passport. A remote server maintains all the passenger information. Electronic Passport information is

•

authenticated and processed with respect to corresponding server information. The passenger is allowed inside the airport only when the information match. By this way passport forgery can be minimized. I have written a tool for analysis of E- passport performance and scalability.

• Performance analysis and scalability measurement for E-Passport Education

05/2011 Bachelor of Technology: Information Technology PSG College of Technology - Coimbatore, India

Interests

• Gardening

• Baking

• Growing herbs, vegetables, or fruits in home gardens.

Contact this candidate