Veydh Gooljar
Software + Electronics Engineer
Machine-Learning + Data Scientist
Email: vgooljar@gmail.com
Location: NYC
Phone: (929) 269-4269
Summary
10 years of experience developing + deploying systems in different ecosystems (AWS, GCP, Linode) using micro-frameworks (Flask, FastAPI, ExpressJS and Go), and larger frameworks (Spring, Django).
I've led large projects within my respective squads involving building data-lakes/pipelines (processing data at the scale of 100s of TBs), mobile-friendly sites (having throughput of millions) and projects incorporating several machine-learning models that I've developed to aid in digitizing data from images.
I've also been involved in Ops rotations; setting up and monitoring Grafana alerting, as well as maintaining AWS codedeploy pipelines and performing rotations for zero-downtime deployment.
Work Experience
Tripadvisor
Software Engineer
Jun 2021 - Present
  • Full stack development across several tech stacks involving Java + GraphQL + React. Python + Angular

    Worked on several large features for menu management and syndication involving:

  • Menu Widget - third party website menu integration
  • Menu Editor - allowing customers to make instantly realized edits
  • Menu Insights - dynamically generates charts and maps for visualizing analytics data
  • Dockerizing microservices to further reduce versioning issues when developing locally
  • Facebook re-integration - Due to Meta's api deprecations; pivoting to utilizing fb posts for menu updates
  • Dish Photos - Extending APIs to allow for ingestion, moderation and addition of photos for dishes

    Hackathon projects:

  • Menu Image to Structured Data: Leveraging OCR + Machine-Learning:

    > Explored Amazon Textract / TesseractOCR
    > Explored text classification models, settling on TfidfVectorizers from XGBoost
    > Produced an independent microservice that can generate structured menu data from a menu image or pdf

  • Menu Item Price Prediction: Based off of Geo and Menu Item name

    > Leveraging data from over 100K menus and their geocoordinates, built a model to predict menu item prices
    > Produced a standalone microservice that can provide a suggested price when given a menu item name and geocoordinates
Prizeout
Senior Software Engineer
Aug 2020 - Jun 2021
  • Extended integrations with Gift Card Aggregator APIs to automate cancellations
  • Developed node microservices in GCS to organize and manage inventory by brands
  • Led data-gathering/web-scraping projects to acquire brand metadata (via Crunchbase)
  • Developed microservice for prioritizing and policing sent emails to avoid spam classification as well as reducing email delivery times
  • Extended Prizeout's Shopify services leveraging Shopify's APIs, providing better UX for Shopify merchants
  • Pioneered a pipeline for Shopify Basic-plan merchants to automate management of gift cards through our system
  • Hackathon: Chrome extension for Prizeout users to have their gift cards auto-filled on the relevant brand's checkout page
  • Developed lead generation tool (Chrome extension + Flask) to aid in finding point-of-contacts for merchants and partners by leveraging LinkedIn contacts.
  • Identified and patched high-level security issues + lower-risk exploits.
  • Revamped UI for several sections of the Merchant and Partner portals (using Less and AngularJS)
  • Created Angular chart components for merchant analytics (D3 charts + ES6)
WiPay
Software Engineer / Consultant
May 2020 - Jun 2020
  • Analyzed system architecture and produced system design documentation
  • Replaced existing Java jPOS server (bank's switchware interfaces for debit card transactions). Following the ISO-8583 protocol a proof-of-concept was implemented with Python's Asyncio
Singleplatform
Software Engineer II
Aug 2019 - Mar 2020
  • Converted processes involved in menu publishing and copying to work asynchronously
  • Reduced clutter/unused services in menu front-end service
  • Extended front-end service to support linkable sub-menus as well as improving SEO enforcing the menu schema expected by Google Structured Data
  • Reduced Race conditions and improved exception handling; better logs in Cloudfront & Sentry
  • Patched security issues that were identified by OWASP ZAP scans
  • QoL Data entry improvements: auto-scrolling to errors + error highlighting and messaging
  • SalesForce integration was made more robust as well as reducing bugs involved in edge-case scenarios for prorated calculations
  • Resolved time localization bugs - improving systems' scheduled behavior (menu publishing)
  • Performed system investigations based on error logs and customer reports, leading to the creation and resolution of Jira tickets
  • Project Leader in refactoring menu scrapers (monitoring + classifying menus across publishers); Utilizing AWS Cloud-formation stacks (Lamba, Glue/ETL, S3) as well as Google BigQuery
Sunny Group of Companies
Lead Electronics Engineer and Software Developer
Feb 2016 - May 2019
  • Reverse Engineered products used for casino machine management
  • Collected extensive data via oscilloscope, processed and deconstructed proprietary protocols to understand and reproduce SAS commands
  • A wireless SoC was then fabricated: ESP8266 microcontroller + Embedded C++ software was developed to interface with slot machines relaying sensor data + events to a central server

    This product/project saved over TTD 1,000,000/year per casino owned and provided the opportunity to compete in a new market.

  • Investigated and leveraged the MQTT protocol for solving the C10K problem which allowed us to easily scale; monitoring 10,000 slot machines on a single server
  • Monitored MQTT data via a web stack: Django + PostgreSQL + Celery + RabbitMQ
  • The GUI for Programming of SoC devices was created with Visual C#. Due to the company's migration from Windows to Ubuntu, this was quickly recreated with Python + QT Framework.
  • A public REST API (token authorization with Django) was implemented to allow our Asset Management system to perform financial analysis based on meter data collected
  • Using Java's Spring Boot + MongoDB + RabbitMQ, an Asset Management system was developed + tested over months of live usage
  • Python scripts were developed that performed system monitoring (through 'top', Apache logs, and Error logs) from selected processes and providing live downtime alerts (via email)
  • Leveraging the created asset management software + newly fabricated electronics, hundreds of work-hours/month was alleviated by optimizing the dispatch of collectors