Nagbhushan Pai

SOFTWARE ENGINEER

Building scalable backend systems and AI-powered products with calm, production-minded engineering.

I design reliable services, data-driven pipelines, and practical ML workflows for recruiters, users, and teams that value systems thinking over surface-level polish.

ResumeView ProjectsContact Me
Backend systems
Distributed systems
AI evaluation workflows
Cloud deployment

Terminal

Terminal

Try: help, about, projects, skills, experience, resume, contact

$ help
Available commands: help, about, projects, skills, experience, resume, contact
10K+
requests/sec
99%
workflow reliability
92%
deepfake accuracy
5
featured projects

About

A portfolio built to read like an engineer, not a resume dump.

The structure emphasizes outcomes, systems, and signals that matter to hiring teams: scope, reliability, technical depth, and evidence of execution.

Currently Building

Active work and learning goals.

A snapshot of what I'm iterating on right now.

A more robust RAG evaluation workflow with cleaner metrics reporting.
A stronger backend systems portfolio narrative with quantified outcomes.
Deeper understanding of distributed systems, observability, and cloud operations.

Experience

Engineering work with real systems pressure.

The experience section leads with backend ownership, workflow reliability, and product impact.

Data Axle

Software Development Engineer (Backend) Intern

Jan 2026 - Jun 2026
Pune, India
Built a standalone DirectMail microservice and reduced reliance on the monolith.
Designed GraphQL APIs with pagination, filtering, and role-based access control.
Implemented resilient Temporal workflows for long-running business processes.
Deployed services on AWS with CloudWatch logging and monitoring.
Improved deployment flexibility by separating critical functionality into independently deployable services.
Reduced debugging turnaround with centralized logs and workflow observability.
PythonDjangoGraphQLGrapheneTemporalAWSDocker

ONJI Softwares Pvt Ltd

Software Engineer Intern

Jun 2025 - Aug 2025
Remote
Built reusable cross-platform UI components for Android and iOS.
Converted Figma designs into production-ready screens.
Implemented secure authentication and navigation flows.
Optimized responsiveness and startup performance.
Delivered a more consistent mobile experience across screen sizes.
Improved maintainability through reusable component architecture.
React NativeExpoJavaScriptREST APIsGit
View Demo

Projects

Recruiter-focused case studies.

Each project card shows the problem, solution, stack, results, and links.

Distributed Rate Limiter

Redis-based traffic control system designed to keep APIs fast while preventing abuse at scale.

Problem: APIs need to enforce limits without introducing high latency or a single point of failure.
Solution: Built a distributed service using Redis and Lua scripting with token bucket and sliding-window strategies, plus configurable policies for different routes.
10,000+ requests/secSub-5ms latencyHorizontal scalability
PythonRedisLuaDocker
GitHubLive demo not availableView case study

Mini Kafka

A Kafka-inspired streaming platform to study partitions, offsets, replay, and consumer group behavior.

Problem: Distributed messaging internals are easier to learn by building the primitives yourself.
Solution: Implemented producer-consumer messaging with topic partitioning, offset management, durable log storage, and concurrent consumers.
Replay supportPersistent logsConcurrent consumers
PythonMultithreadingNetworkingFile Storage
GitHubLive demo not availableView case study

RAG Evaluation System

A benchmark framework for comparing LLMs on scripture-based question answering with retrieval and quantitative evaluation.

Problem: LLMs often respond confidently without enough grounding for culturally specific or knowledge-heavy prompts.
Solution: Built a retrieval pipeline with Sentence Transformers and FAISS, then evaluated multiple models using BLEU, ROUGE-L, and BERTScore.
Automated benchmarkingGrounded retrievalQuantitative reports
PythonFAISSSentence TransformersHugging FaceLLM APIs
GitHubLive demo not availableView case study

Audio-Video Deepfake Detection System

Multimodal AI system for identifying manipulated videos with spatial and temporal features.

Problem: Manipulated media needs fast, explainable detection for practical real-world use.
Solution: Combined CNN and Vision Transformer features in a real-time inference pipeline with preprocessing and REST deployment.
92% classification accuracyReal-time inferenceExplainable scoring
PyTorchOpenCVFlaskVision TransformersCNNs
GitHubLive demo not availableView case study

Blog

Three short posts that signal real technical judgment.

These posts are intentionally concise, recruiter-friendly, and tied to the projects in the portfolio.

2026-06-24

Building Kafka From Scratch

A short breakdown of the ideas behind partitions, offsets, replay, and durability.

Read post

2026-06-24

GraphQL vs REST at Scale

A practical comparison of API design tradeoffs when traffic, teams, and data access patterns grow.

Read post

2026-06-24

How We Built a Deepfake Detector

A concise look at multimodal detection, preprocessing, and explainable scoring.

Read post

Timeline

A concise professional arc.

This is the version recruiters can scan quickly to understand growth across leadership, research, and shipping work.

2026

Current work

Shipping portfolio and engineering polish

Extending backend, AI, and frontend systems while keeping the site recruiter-focused.

2026

Data Axle Internship

Backend systems and workflow orchestration

Built GraphQL services, Temporal workflows, and AWS-backed infrastructure.

2025

Deepfake Project

Multimodal detection system

Developed an AI pipeline for real-time manipulated media detection with measurable accuracy.

2024

SIH Finalist

Smart India Hackathon 2024 finalist

Built an AI-powered multimodal solution and presented it to evaluators.

2024

NCC

Sergeant, C Certificate

Led teams, coordinated activities, and built discipline through cadet responsibilities.

2024

Dronaid

Electronics team contributor

Worked on sensor integration, embedded systems, and flight testing support.

Skills

Core capabilities

A compact view of the engineering areas this portfolio is built to communicate.

Backend EngineeringDistributed SystemsMachine LearningCloud InfrastructureGraphQL APIsWorkflow OrchestrationData PipelinesTechnical Writing

Achievements

Quantified signals

These numbers help the page read like evidence, not decoration.

Smart India Hackathon 2024 Finalist
NCC Sergeant with C Certificate
IIT BHU research project contributor
92% deepfake detection accuracy
10K+ requests/sec rate-limiter design

Contact

Open to backend, AI, and systems-focused opportunities.

Keep the contact block short and action-oriented so recruiters can reach out quickly.

Email
nagbhushanpai707@gmail.com
GitHub
github.com/NagbhushanPai
LinkedIn
linkedin.com/in/nagbhushan-pai