Satish Pokala

Full Stack Engineer | American Express GBT (July 2022 to present)

• I majorly developed the backend infrastructure for an internal performance monitoring tool called Quest, which saves about a million dollars annually for the company.
• Quest helps monitor employee performance by bringing numerous KPI data from various data sources together and presenting it in simple, consumable formats, such as charts and tables.
• I take a key role in architecting and implementing the authentication and backend solutions for this system by leveraging Okta SSO, Python, and FastAPI for secure, robust API development. The API integrates with PostgreSQL for efficient data management, AWS services such as S3 for scalable storage, and EC2 for running various automation tasks.
• Built the prototype of the whole authentication system, which utilizes the company's SSO(Single-Sign-On) provider, Okta. I have learned various security standards during my research phase, such as OAuth 2.0, OIDC, and SAML. I was given 2 weeks for this task, but I built it in just three days, making an excellent first impression among the team members.
• With that impression, I was challenged yet again to build more prototypes. One is to build the file storage solution using AWS S3. This system is secure and tolerant of file upload exploits as I had designed it to verify actual file contents without just relying on the extensions as they are easy to manipulate. Another prototype was to build the secrets management solution using AWS Secrets Manager eliminating the need for developers to store critical secrets like API keys, DB passwords, and other shared credentials.
• The prototypes that I built, made building our actual application so simple as we knew all the challenges beforehand and solved them in the best way possible.
• As time passed, we had to expand to multiple countries and regions. Some countries have a specific set of privacy rules in terms of data visibility. These factors lead to increased complexity of authorization, data visibility, and processing. This lead to degraded performance for specific features. One such feature relied on an endpoint that declined in performance as the user base grew. In just a few weeks we have timeouts in production affecting user experience. I addressed the root cause which is unoptimized queries and database schema for large data sets.
• I optimized the queries by reducing joins, avoiding unwanted data for processing, and utilizing CTEs wherever possible to process data in small pieces. I researched various optimization techniques like indexing and added a Btree index on a combination of columns that significantly boosted performance by bringing the response times from 30s to just under 1s.

Engineer | American Express Travel (July 2021 to June 2022)

• Supported Amex loyalty applications called Star Service Layer (SSL), and Solitaire. SSL acts as an interface for another application called Star. SSL makes it easy for Star to consume events from multiple systems.
• Developed Python and Java automation scripts that issue rewards for select customers, provided by the business. This significantly improved our team productivity and eliminated human errors while providing rewards to the customers.
• Performed load and endurance testing using tools like JMeter and Postman.
• Deployed fixes for security vulnerabilities and managed certificate renewals.