Pawankumar B

Full-stack software engineer — cloud and backend.

Bengaluru, India · open to remote SDE2 / full-stack roles.

Full-stack engineer, but I lean backend. I'm at home designing APIs and data models, owning the deploy pipeline, and stepping into the frontend when the team needs it — not because I love CSS, but because shipped beats pretty.

Currently building systems software at Ninestars — Intel SGX integrations, secure enclaves, and a custom Linux ISO that bootstraps Kubernetes clusters with zero touch. Before that I led a small team at a VIT-incubated startup and co-authored a book chapter on steganography in cloud environments, which is a sentence I never expected to write.

Now

Production work on SGX runtime integration and zero-touch K8s provisioning. On the side, building a Django + Celery + Redis distributed task scheduler — to go deeper on async workers, retry semantics, and what to do when a queue is on fire.

Experience

Software Engineer — OS & Security · Ninestars Information Technologies

Jan 2025 — Present

Bengaluru

  • — Integrating Intel SGX and TME into core runtime to protect sensitive data in execution and in memory.
  • — Designing hardened Kubernetes deployments with RBAC, pod security policies, Helm, and kubeadm.
  • — Built a custom Linux ISO with master/worker role assignment for zero-touch Kubernetes provisioning.
  • — Embedded Ansible inside the ISO for centralized, reproducible cluster configuration across nodes.

Lead Developer · Valsco Tech (VIT-incubated)

Dec 2023 — Mar 2024

Vellore

  • — Led a team of 9 across two client projects — task delegation, code review, release coordination.
  • — Owned architecture, sprint planning, and end-to-end delivery for both projects.

Selected projects

Environmental Telemetry RAG

Active

Python · AWS · Pinecone · Node.js · REST

Real-time retrieval pipeline for state pollution-control bodies — ingests environmental sensor telemetry, indexes semantic context in Pinecone, and exposes insights through a REST API with a companion mobile chat client. Joint research between Middlesex University (London) and Vellore Institute of Technology.

Distributed Task Scheduler

Building

Django · Celery · Redis

A small distributed job runner around Django, Celery, and Redis — retry semantics, idempotency, and a dashboard for the times you actually need to look at a queue at 2 AM.

The rest (3)

AI Interview Platform

Active

Django · React · Ollama · PostgreSQL

Self-hosted platform that runs structured technical interviews and ships transcripts with rubric-based scoring. Django REST backend, React frontend, Ollama for local LLM inference — candidate data never leaves the host, which is the entire point.

GostWatch — Honey Tokens on AWS

Shipped

AWS · Terraform · Python · Slack

Terraform-deployed honey tokens on AWS that lure unauthorized access and fire Slack alerts in under a minute. Integrated with GitHub and GitGuardian to flag leaked credentials before someone else does.

DanceHolic — AI Dance Instructor

Shipped

TensorFlow · React · Python

React app using TensorFlow pose detection to track dance movements in real time. Tuned the model and smoothing pipeline to ~95% landmark accuracy. My dancing did not improve.

Education & recognition

M.Tech (Integrated), Software Engineering · Vellore Institute of Technology

CGPA 8.23

2020 — 2025

Thesis · M.Tech, VIT

Steganography in Cloud Computing Environments

Published as Chapter 10 in Enhancing Steganography Through Deep Learning Approaches, IGI Global, 2025. Co-authored with Dr. Siva Rama Krishnan (Vellore Institute of Technology) and Nithyashree R.

ISBN 9798369322239

Certifications

AWS Certified Cloud Practitioner · Applied Data Science (Google Developers)

Skills

Roughly, the stack I'd be comfortable getting paged about.

Languages
Python, TypeScript, JavaScript, SQL, Bash
Backend
Django, FastAPI, Node.js, REST APIs, async workers
Frontend
React, Next.js, Tailwind CSS — enough to ship without slowing the team down
Data
PostgreSQL, MongoDB, Redis, Pinecone
Cloud
AWS, Kubernetes, Docker, Terraform, Ansible, Helm, GitHub Actions
Systems
Linux, Intel SGX, RBAC, X.509, cryptography fundamentals — the layer most devs route around
ML / AI
RAG pipelines, LLM integration, computer vision — comfortable wiring models into production, not auditioning to be a researcher
Practices
API design, system design, observability, IaC, CI/CD, code review

Contact

Fastest way in is email — I usually reply inside a day on weekdays. I'm also on GitHub and LinkedIn. No cold pitches for SEO services, please.