SWAYOM.AIv2.4
About

The engineer behind the systems

I operate at the intersection of three worlds — the offensive mindset of a security researcher, the analytical precision of a data scientist, and the physical intuition of an embedded systems engineer.

01 / Philosophy

Design principles that govern everything

⬡

Minimal Complexity

Every abstraction must earn its existence. Clean systems beat clever ones.

◈

Adversarial Thinking

Build for the attacker's perspective. Security is a design property, not an afterthought.

◉

Edge First

Compute at the source. Latency is the enemy of real-time intelligence.

⊕

Data Honesty

Models don't lie. But datasets do. Rigorous validation over impressive metrics.

02 / Journey

The path that shaped the stack

2024

Neural Systems Engineer

Independent Research & Freelance

Designing ML-powered security tools, IoT sensor networks, and full-stack platforms for research-stage startups.

2023

B.Tech — Computer Science

Specialization: AI & Cyber Security

Deep dive into adversarial ML, network security, embedded systems design, and large-scale data pipelines.

2022

IoT Deployment — Himalayan Project

Rural Tech Initiative

Deployed 40+ LoRaWAN sensor nodes for water quality monitoring across remote Himalayan villages. Edge AI processing.

2021

First Penetration Test

Bug Bounty — HackerOne

Discovered a critical authentication bypass in a fintech platform. Began systematic study of offensive security.

03 / Stack

Tools of the trade

Languages
PythonTypeScriptC/C++BashSQL
AI / ML
TensorFlowPyTorchScikit-learnHugging FaceLangChain
Security
Kali LinuxWiresharkMetasploitBurp SuiteNmap
IoT
ArduinoESP32Raspberry PiMQTTLoRaWAN
Infrastructure
DockerKubernetesLinuxAWSNginx
Databases
MongoDBPostgreSQLInfluxDBRedisMySQL

“The most dangerous engineer is the one who understands both how to build a system and how to break it.”

— Swayom Sharma