Backend engineer for high-throughput systems

I build APIs, pipelines, and services that stay calm under load.

Senior backend engineer with 5+ years in ad-tech, recommendation systems, serving-path migrations, and performance-focused backend platforms.

View engineering stories Contact me

production thinking

REST APIs ~1K QPS

Kafka flows serving path

Ranking recommendations

Low latency -250 ms

Business lift $120K/mo

Scale handled 3M impressions/day

Latency improved 250 ms faster

Throughput scaled 10-12 QPS to 50 QPS

Experiment impact 25% to 2% low traffic

Selected engineering stories

Proof without pretending every project needs a landing page.

Until the side-project catalog is ready, the portfolio can still show the problems I solve: performance, throughput, recommendation quality, and production migration work.

Serving path migration Backend systems

Moved traffic through faster recommendation flows without hurting delivery.

Worked on backend experiments and serving-path changes for ad-tech systems where latency, impression quality, and delivery stability all mattered.

Latency: -250 ms
Traffic: ~1K QPS
Domain: ad-tech

Recommendation throughput Data pipelines

Scaled keyword recommendation throughput for higher-volume workflows.

Improved pipeline capacity while keeping the system practical for production constraints and downstream serving needs.

Before: 10-12 QPS
After: 50 QPS

Experiment quality Ranking systems

Improved impression outcomes across large-scale recommendation experiments.

Focused on measurable backend impact: better traffic allocation, stronger recommendation performance, and clearer experiment feedback loops.

Lift: 30%
Scale: 3M/day

How I think

Backend strengths shown as decisions, not badges.

Design APIs around workflows

Start with the user or service flow, then shape endpoints, payloads, validation, and failure states around the real operation.

Measure before optimizing

Treat latency, QPS, queue depth, and experiment outcomes as design inputs, not post-release trivia.

Prefer boring reliability

Use clear data models, idempotent jobs, safe migrations, and observability before chasing novelty.

Stack

Tools I use to ship backend systems.

Backend

Java, Python, Spring Boot, Node.js, REST APIs

Data and infra

Kafka, Aerospike, Redis, MySQL, MongoDB, Milvus

AI-adjacent systems

LLMs, NLP, vector embeddings, Scikit-Learn, Pandas

Delivery

Prometheus, Grafana, Jenkins, Docker, CI/CD

Beyond backend

Close to the product side of technology, too.

Outside backend engineering, I have been part of invite-only product feedback and beta programs for flagship consumer technology. It adds a useful product-quality lens to how I think about software.

Global product feedback

Represented India at OnePlus Global Open Ears Forum in Italy.

Invited to share structured feedback on OxygenOS with OnePlus, bringing a user, product, and software-quality perspective to the discussion.

Beta testing

Tested OnePlus 13 Series and OxygenOS flagship beta rollouts.

Participated in beta programs for OnePlus 13 Series devices and OxygenOS 15 and 16 rollouts, reporting product behavior and release feedback.

Available for backend roles

Let’s talk about APIs, scale, and systems that need careful hands.

morje.amey@gmail.com GitHub LinkedIn