Senior Associate, Software Engineer
I am actively developing & managing micro-services. Currently I am Working in both frontnend and backend as well as maintaining cloud resources.
I am a core member of ACM (astro.com.my) under Digital Ecosystem (D5). which manages Homepage, ACM Shop, ACM Rewards & Promotions, ACM Content guide.
With 75% market penetration Astro is the largest payTV network in Malaysia. D5 manages the digital footprint of the company.
ACM is the frontline for entire Astro. Hence making sure our resources and services are secure and performant is a top priority and under my jurisdiction.
Challenges include making sure we can accommodate campaign traffic, allow users to redeem during flash sale.
I ensured 100% deliveries on time and without any post deployment hotfix. In ACM Shop, the lead capture was improved by 48.66% mainly by
designing fallback approaches and effective monitoring. Area coverage was a major concern for lead capture, I helped design the system in a way
so that users find it more flexible. It helped reduce the churn rate by 5%.
Each API latency must be < 500 ms for seamless customer experience in a mobile first world. I ensured the SLA is met by monitoring the latency
across services and caching the responses. Warming up lmabda is another approach for ensuring serverless services are active.
Frontend performance is of paramount importance. Instead of bundling entire components, loading by chuks helpe reduced the TTI significantly.
It cut the TTI and improved the score by 2 sec.
Frontend errors are hard to catch and resolve. I helped ensure 65% such errors are reported and resolved within 2 weeks. Specially the mobile devices with older
browsers as well as in app browsers cause incompatibility issues which is hard to track. So I meticulously went through each report
and prioratize accordingly. While adding polyfills can solve it but it causes performance penalty. So I had to carefully balance the
tradeoffs.
Frontend errors are hard to catch and resolve. I helped ensure 65% such errors are reported and resolved within 2 weeks. Specially the mobile devices with older
browsers as well as in app browsers cause incompatibility issues which is hard to track. So I meticulously went through each report
and prioratize accordingly. While adding polyfills can solve it but it causes performance penalty. So I had to carefully balance the
tradeoffs.
21% cost reduction was possible to due scheduled downsizing of non-critical resources. Also properly collaborating with the resource owner, I managed to
decomission redundant services / helped migrate them to a shared resource model instead of dedicated one.
Existing and legacy code coverage was improved to 99.63% including the unit tests and integration tests.
Debugging a service through multiple lambda logs is quite time consuming. So I helped wrote a tool to span timeline of a request across multiple
lambda functions. Astro-Dissection
I co-ordinate backend sessions within Digital ecosystem. We present our architecture and discuss on the tradeoffs and how to improve the reliability and security of the services.
Knowledge gathered:
- Typescript
- Nodejs
- Serverless
- React
- Redux
- AWS
Senior Software Engineer
Prime responsibilities include architecting solution and developing them.
I am a core member of Smartmed Healthcare Solution which manages critical patient's data.
I have helped develop doctor's collaboration platform with flexible access rule engine. Gradually I instigiated the process to profile and refactor the legacy codes with test coverage.
Currently the web performance gain is 3 times compared to initial state.
Apart from developing optimized API for mobile clients, I developed the Pubish subscribe component and helped to deploy the distributed infrastructure.
Knowledge sharing is an integral part of my professionalism. I mentor junior developers and take technical sessions regularly.
Knowledge gathered:
- python
- django
- Mysql
- Rabbitmq
- bootstrap
- highchart
- jquery
- NodeJs
- redis
- socket.io
- REST API
Software Engineer
Key responsibilities were developing backend services and API. I have automated the deployment with ansible and docker. Bundle scanning process was a crucial part of the system.
I designed it with proper software and database locks. This improved the throughput and operational speed. Error ratio went from 30% to less than 5%.
I have monitored the deployed services. Mostly observing key memory usage over time to identify memory leaks. I observed other metrics like CPU usage and Mysql handler threads to determine
any discrepency in not freeing hardware resources by application process.
Knowledge gathered:
- NodeJs
- sequelize
- Mysql
- Redis
- REST API
