Tapan Parmar👋
Versatile Full Stack Developer with expertise in Python backend development, frontend technologies, and cloud services
Client For:
Vectoscalar
Services:
Python, Flask, FastAPI, GoLang , React JS, React Native , ELK, Grafana, Pagerduty, New Relic, AWS SQS, SNS, Lambda, ECR, EC2, Load balancer, Cloudwatch, Batch ,MongoDB, Firestore , PostgreSQL, MySQL Docker, Kubernetes
https://coinswitch.co/Overview
CoinSwitch is now India's largest crypto app with over 18 million users and more than 500 employees. We're backed by some of the world's leading investors including a16z, Tiger Global and Sequoia Capital India. The cornerstones of coinswitch's culture are customer obsession and an employee-first approach. They're building a platform that democratizes investments for everyone making them simple, safe and accessible to every Indian.
Wireframing and Prototyping: Developed and enhanced the rewards, currency listing, and crypto learning backend sections for CoinSwitch's cryptocurrency trading app using Flask and Python scripting. Designed and implemented REST/RESTful APIs to support major app features such as rewards and Crypto bytes. Improved user interaction by developing the initial version of Crypto bytes backend, one of the most successful experiments in the team's history. Increased code coverage and wrote unit tests for rewards and currencies sections after integrating Sonar and finding low code coverage. Utilized Redis caching and SQL query optimization techniques to maintain low latency and high throughput. Designed a comprehensive solution for all coin listing API performance using Firebase Realtime database, SQS-based messaging queue, and Postgresql to deliver high RPS REST APIs to mobile apps, the main website, and third-party API consumers. Created backend APIs for top gainers and top losers features, which were integrated with currencies API and had an average of ~1M calls per hour. Contributed to optimizing the codebase and setting up standard practices as part of the peer review group. Participated in performance and security meetings to improve monitoring measures. Created dashboards in New Relic and Grafana to monitor matrices for various high-impacting backend areas
Challenges
High volume trading:
- Challenge: Coinswitch is a crypto trading platform with high volume trading and ~1M+ DAU as a baseline capacity
Monolith and microservice architecture :
- Challenge: Understand how Monolith + Microservices architecture combination works and serves APIs to internal as well as external entities
Caching layer:
- Challenge: Caching layer was quite complicated and specific
Results/Conclusion:
Independent AWS lambda functions for monitoring and reporting.I was actively involved in brainstorming sessions for decision making regarding architecture choices to help team drafting tech solutioning docs. Helped release two significant features, which allowed the client to generate more insights about user behaviours and enhance user engagement. Cryptobytes helped the team get insights into user reading interests that correlate highly with purchase patterns. Putting knowledge-sharing features and polls in the next version could involve users more and improve the average time spent on the app. Alignment and enhancement of currency listing endpoints to a single endpoint helped us manage a single point of interaction and robust access patterns implementation. Multiple reporting dashboards with combination of New relic, Grafana and Pagerduty integration helped them achieve overall system monitoring.
