Seeking: Embedded Systems & Security Engineer

Hi, I'm Mangesh Tiwari

Electronics & Telecommunication engineer with hands-on experience building real-time control systems (200Hz PID), computer vision pipelines (96% accuracy), and network security tools. Comfortable working across the full stack — from bare-metal firmware to ML model deployment.

scroll

Engineer who builds
things that move.

Fresh Electronics & Telecommunication graduate from St. Francis Institute of Technology, Mumbai (2026). Built a self-balancing agricultural rover with custom Kalman filtering and 200Hz PID control as my final-year project.

I focus on problems where software meets hardware — real-time control systems, computer vision pipelines, and network security tooling. Competed in Flipkart Grid 6.0 (Computer Vision track) and IIT Bombay Techfest (Line Follower).

Currently expanding into cybersecurity through hands-on labs on TryHackMe and Google Cloud certifications.

96%
ML Accuracy
200Hz
Real-time PID
5
Production Systems
3+
Domains
🎓
B.E. Electronics & Telecommunication
St. Francis Institute of Technology, Mumbai
2022 — 2026  ·  CGPA: 7.0

Engineering solutions.

Each project started with a real problem and ended with working hardware or software — here's what I built and why.

Flip-Check freshness detection dashboard — Flipkart Grid 6.0
Flipkart Grid 6.0

Flip-Check — Freshness Detection System

Problem → Grocery supply chains need a fast, automated way to check whether fruits and vegetables are still fresh — manual inspection is slow and inconsistent.

Solution → Built a YOLO-based computer vision system that classifies produce freshness with 96% accuracy. Integrated barcode scanning (PyZbar), OCR (Tesseract), and object counting into a single real-time Streamlit dashboard — one screen showing freshness scores, product info, and inventory count. Trained on a custom dataset using Roboflow SDK for annotation and augmentation.

YOLO Python OpenCV Streamlit PyZbar Tesseract OCR Roboflow SDK
Network Intrusion Detection System visualization

Network Intrusion Detection System

Problem → Network admins can't manually monitor every packet — they need automated tools that catch port scans, DDoS floods, and unauthorized access in real time.

Solution → Trained Random Forest and SVM classifiers on network traffic dumps to flag malicious patterns. The system catches port scans, DDoS attempts, and suspicious login behaviour in simulated environments — filtering real threats from normal traffic without flooding the admin with false alarms.

Python Scikit-learn Pandas Network Security Anomaly Detection
Face gender age detection system with ESP32 and Raspberry Pi Zero
Hardware + ML

Face, Gender & Age Detection

Problem → Surveillance cameras generate hours of footage but can't extract any useful info — you still need a human to identify age, gender, or count people in real time.

Solution → Built a standalone detection unit using an ESP32 and Raspberry Pi Zero with a camera module. The Pi runs OpenCV's DNN module with pre-trained Caffe models for face detection, then pipes frames through age/gender classifiers (tested Logistic Regression, SVM, and a small neural net). The ESP32 handles I/O, triggers, and communication with external displays. The whole setup runs headless — plug in power and it starts detecting.

ESP32 Raspberry Pi Zero Python OpenCV Caffe DNN SVM Camera Module
Real-time FIR audio filter on Basys-3 FPGA
FPGA / RTL

Real-Time Audio FIR Filter — FPGA

Problem → Software audio filters introduce latency. For real-time processing you need the filter running in hardware — but designing FIR engines on an FPGA means writing your own multiply-accumulate pipeline from scratch.

Solution → Implemented a 45/89-tap FIR filter in SystemVerilog on a Basys-3 FPGA. Designed four switchable filters — low-pass (1 kHz), high-pass (2 kHz), bandpass (1–4 kHz), and moving average — with hardcoded 16-bit coefficients. Audio comes in through an I2S (AXI-Stream) interface, gets convolved in real time, and goes out with zero software latency. Used MATLAB for coefficient generation and wrote SystemVerilog testbenches for verification.

SystemVerilog FPGA (Basys-3) FIR Filter I2S / AXI-Stream MATLAB DSP Vivado
Line follower robot for IIT Bombay Techfest
IIT Bombay Fest

Line Follower Bot

Problem → A bot needs to follow a black line on a white track at speed, handle sharp turns, and not lose the path — all autonomously in a timed competition.

Solution → Designed the sensor layout with an IR sensor array at the front and wrote a PID control loop to keep the bot centred on the line even through tight curves. Tuned the P, I, and D gains for the fastest stable run. Competed at IIT Bombay's annual tech fest against teams from engineering colleges across India.

Embedded C IR Sensors PID Control Motor Drivers Arduino

Tools of the trade.

Embedded & Control Systems

ESP32 Arduino Raspberry Pi PID Control Kalman Filtering Sensor Fusion Stepper Motor Control DM542 / TB6600 Drivers NEMA Motors MPU6886 IMU
💻

Programming

C C++ Python Embedded C MATLAB
🔒

Security Fundamentals

Network Security Penetration Testing (TryHackMe) Threat Analysis Wireshark
🧠

ML & Computer Vision

YOLO OpenCV Scikit-learn Streamlit Roboflow SDK Tesseract OCR Anomaly Detection
🛠️

Tools & Platforms

Linux Git & GitHub Vivado (FPGA) Google Cloud Arduino IDE PlatformIO Wireshark

Where I've worked.

Technical Intern
CyberStance
May 2025 — Jul 2025
  • Studied .NET and binary malware samples; documented behavioural indicators for team reference
  • Researched EDR tools and compiled a comparison of open-source vs commercial solutions
  • Explored how AI is applied in SOC workflows — automated triage and alert correlation
  • Wrote weekly write-ups on threat detection concepts and response procedures
🏴‍☠️
Junior Penetration Tester (PT1)
TryHackMe
2025 — Hands-on labs: network attacks, Active Directory exploitation, web app security
☁️
Google Cloud Cybersecurity
Google
2025 — IAM, encryption, compliance, cloud workload security
🛡️
Foundation of Ethical Hacking
Udacity (Accenture)
2025 — Penetration testing fundamentals, vulnerability assessment, risk management

Let's connect.

I'm seeking full-time roles in embedded systems, robotics, or cybersecurity engineering. Open to relocation.