Apr 2019 – Aug 2021 · 2 yrs 4 mos · New Delhi, Delhi, India

• In my third year at DailyObjects, the first task was to manage the AWS costs. This was reduced by 45%, with the following optimizations as applicable:
• 1. EC2 On Demand -> Spot + Scheduling
• 2. Docker Container based Multiplexing*
• 3. Classic Load Balancers -> Application Load Balancers
• 4. Removal of redundant unused resources
• 5. Use of S3 Storage Classes
• After that the team and I worked upon streamlining the deployment processes, testing processes and disaster alerting/management strategies. We set up SNS and Jenkins Polling Pipelines to continuously monitor our services.
• Working with the marketing team, design team and founders to organize processes, support decisions with data and provide technology to gather data was the prime task in early 2020.

Jan 2018 – Mar 2019 · 1 yr 2 mos · New Delhi, Delhi, India

• In my second year, along with working on the android app I took up work related to python development. I worked on setting up the user events logging system based on RabbitMQ, MongoDB, Spring Boot, Uniqush ( for notifications ) and Redis.
• Halfway into this year, the management changed and the new tech lead introduced a great open culture into the company's tech. This was a turning point as I was introduced to AWS and the DevOps work. I had extensively worked on bash and back-end while developing "Decser" therefore this transition was relatively smooth.
• After this I was assigned to a CNN based software, for detecting designs over products to help the operations team. The project could successfully classify designs, however speed was an issue. The code was an adaptation from a research paper, modified to suit our needs.
• Last project for this year was an image stitching software with back-end in Python and front-end in Kotlin ( TornadoFx). The back-end was an implementation of the bin packing algorithm. The challenge in front-end was to work efficiently with super large bitmaps.

Feb 2017 – Jan 2018 · 11 mos · New Delhi, Delhi, India

• This was my first job in the industry. Here is when I experienced what it felt for your app to be used by over 100K users. I learnt about the feature development life cycle and about how to independently handle a complete app for a particular platform in the absence of a lead. Knowledge (and best practices) were predominantly gathered from conference talks and open source projects like “Mozilla Klar” etc.
• Achievements:
• 1. Animations + Shared Transitions,
• 2. Payment Method Integrations,
• 3. MVVM Architecture
• 4. Crash-free rate to over 97%.
• 5. Memory Leak Removal
• 6. 60fps performance optimizations

May 2018 – Dec 2019 · 1 yr 7 mos

• User research and the Complete launch lifecycle with weekend coding sprints. This android app uses a tensorflow to segregate memes from important photos in your WhatsApp media via transfer learning. Simple but useful.

Jun 2015 – Dec 2016 · 1 yr 6 mos · Noida, Uttar Pradesh, India

• Ideation to implementation of Little Servers on Raspberry Pi, for home consumers. They could give the user a social network, a download manager, and a remote light switch without a centralised server. Everything was based on SSH-ing, port forwarding, xmpp and http. The pod was supported by an android app.
• Legacy Code here :
• https://github.com/freevnshns