Software Engineer
Backend Engineering • Cloud Infrastructure • Distributed Systems
Designing reliable backend services, containerized workflows, and cloud-native architectures with a focus on performance and automation.
About
I'm a software engineer focused on building reliable, cloud-native systems that scale. My work revolves around containerized architectures, distributed systems, and automating everything from deployment workflows to complex ML pipelines. I enjoy designing backend services with Python, orchestrating workloads across Kubernetes, and shaping infrastructure using Terraform and modern DevOps tooling.
I've built and deployed systems across domains like IoT, security automation, and large-scale ML workflows — from real-time telemetry pipelines to autonomous incident response agents. I care deeply about observability, performance, and writing code that is maintainable, predictable, and production-ready.
I'm constantly exploring better ways to design resilient systems, contribute to open-source tooling, and push the boundaries of cloud engineering. Whether it's improving monitoring stacks, optimizing event-driven architectures, or shipping end-to-end platform features, I aim to build systems that are elegant, scalable, and operationally sound.
Featured Projects

Containerized ML Orchestration Pipeline
Production-grade ML orchestration system automating end-to-end Geneformer fine-tuning workflows. Features GPU-enabled scheduling, automated versioning, and complete observability stack (Prometheus, Grafana, cAdvisor) for real-time monitoring. Migrated to AWS EKS using KubernetesPodOperator and Helm for scalable, cloud-native execution.

IoT Fleet Management System
Real-time IoT device management system with MQTT messaging, FastAPI backend, and telemetry dashboard for monitoring distributed vehicle fleets. Features vehicle history tracking, live map visualization with route markers, speed/time graphs, and maintenance ticket management.

IncidentResponder.AI
Autonomous cloud SOC and threat response simulator using containerized AI agents to ingest GuardDuty findings, perform log forensics, apply MITRE ATT&CK attribution, and execute rule-based remediation workflows. Features multi-agent decision flows, secure config management, and audit-grade logging.

CryptoPulse
Real-time cryptocurrency analytics platform with async FastAPI backend delivering live market prices, time-series data, and DeFi TVL metrics. Integrates CoinGecko and DeFiLlama APIs with WebSockets for low-latency updates. Features Redis caching, PostgreSQL storage, and modular OOP backend design.
Skills
Programming
Backend & APIs
Cloud
Databases & Messaging
Observability & DevOps
AI/ML & Automation
Development Practices
Experience
Software Engineering Intern
Nokia • Bangalore, India
- Automated monitoring & visual dashboards to highlight performance degradation, reducing investigation time by 80%.
- Surfaced operational insights to improve energy savings transparency with multi-vendor support & displayed energy/cost impact for every cell across the RAN, Supported with comprehensive testing and CI.
- Developed a modular Python-based configuration validation framework that cross-referenced INI files with parameter metadata parsed from technical PDFs, identifying misconfigurations and missing settings.
- Supported feature rollout by validating system behavior, ensuring safe deployment & improving operational visibility.
- Worked in Agile sprints, contributed to code reviews, demonstrated features & supported test automation efforts.
Publications
Research work published in IEEE conferences, focusing on machine learning applications and automated systems.
xG-Based Football Scouting System using ML Techniques
A data-driven football scouting framework that leverages machine learning-based xG modeling to evaluate player performance and tactical efficiency. The system analyzes match events, player behavior, and chance quality to derive objective scouting metrics. Demonstrates how ML can modernize talent identification and tactical analysis within professional football environments.
Automated Threat Detection in the Dark Web: A Multi-Model NLP Approach
A multi-model NLP system for automated dark web threat detection, integrating transformer architectures (BERT, DistilBERT, DarkBERT) fine-tuned on DNRTI, Agora, and CoDA datasets. Achieved >95% accuracy, deployed with Firebase for scalable classification, visualization, and automated threat reporting.
Let's Connect
I'm always interested in hearing about new opportunities, collaborations, or just connecting with fellow engineers. Feel free to reach out!