I take products from nothing to shipped: user research and UI through to the AI pipelines and the cloud they run on. A content platform I built alone end to end, a CPA-optimization decision engine that has to explain every call it makes, and the AWS and GCP estate underneath all of it.
- Now
- SDE2, Growthz. Since Jan 2025.
- Cloud
- AWS and GCP. Pulumi, Cloud Run, ECS Fargate.
- Focus
- Decision engines, AI pipelines, ad platform data.
Selected work
SDE2 · since Jan 2025
Growthz builds marketing automation: URL scraping, AI creative generation, campaign management and analytics across Meta, Reddit, LinkedIn, ASA, Google Ads, DV360 and AppsFlyer. Everything below is work I own or co-own there.
The one I own outright. Every layer was mine: user research, product vision, UI/UX, frontend, backend, the AI pipelines, cloud deployment, and the post launch loop, with real users, testimonials and feedback folded straight back into the roadmap.
Give it a URL and keywords; it returns platform native content for six channels (Reddit, Instagram, Amazon A+, LinkedIn, Twitter and blogs) with one click publish. Not one draft reformatted six ways: each channel gets copy written to how that channel actually reads.
Visit rightcontent.ai ↗How I work
Education
SRM Institute of Science and Technology
B.Tech, Computer Science and Engineering
Sept 2020 to Jun 2024 · 8.5 CGPA
I like the problems where the correct answer isn't obvious and being wrong is expensive. Thrive is the clearest example. A recommendation that moves someone's ad spend has to explain itself, expire on time, and never silently contradict another lever. So it carries an audit trail, a conflict resolver, and a deliberate path back to deterministic rules when the model doesn't have enough data to be trusted.
The same instinct shows up in the infrastructure. Reading Meta's rate limit headers and applying backpressure instead of retrying into a ban. Idempotent daily aggregates so a re-run can't double count. Running the GCP and AWS estate side by side, and an in-house NAT that took 30% off the AWS billwithout touching uptime. None of it is clever for its own sake. It's what keeps the thing running when nobody's watching.
Experience
Own Thrive end to end and sit second on Wildcard's commit list. Designed the video ad pipeline, and own the GCP and AWS infrastructure: Cloud Run, Cloud Tasks and Cloud SQL alongside Pulumi provisioned ECS Fargate, RDS and S3. Cut AWS spend roughly 30% at 99% uptime. Also client side, scoping and running accounts directly.
Migrated the Supabase backend to Microsoft Azure with Prisma, and contributed to a GenAI document generation feature that put LLM pipelines into the product workflow.
Built a quiz creator and college predictor in Next.js and Tailwind, with secure paginated CRUD endpoints behind them.
Built a cross platform Electron.js desktop app and containerised the backend in Docker for Windows and Linux.