0%
OPEN FOR ROLES// STOCKHOLM, SE

HARSHIT PODDAR

> |

View Architectures Extract Weights (CV)
SCROLL
// 01. INTRO

About Me

I build intelligent, scalable systems. As a Machine Learning Engineer and M.Sc. student at KTH Royal Institute of Technology, I specialize in transforming complex multimodal AI and computer vision research into deployable, high-performance architectures.

My technical foundation is built on hands-on R&D across top-tier aerospace, defense, and academic labs. Whether it's architecting RAG pipelines, automating scene graphs for autonomous agents etc. I focus on solving data-heavy challenges with precision engineering.

Actively seeking Master's Thesis collaborations and Full-Time engineering roles to build the next generation of AI solutions.

0+

Publications

0

Patents Filed

0+

Lab & Industry Roles

0

Awards Won

// 02. KNOWLEDGE GRAPH

Academic Base

2025 - 2027 (Ongoing)

M.Sc. in Machine Learning

KTH Royal Institute of Technology | Stockholm, SE

2021 - 2025

B.Tech in Electronics & Communication

Vellore Institute of Technology | Vellore, IN

CGPA: 4.4/5

// 03. OPTIMIZATION LOGS

Career Trajectory

Student Researcher

Robotics Perception and Learning (RPL) Lab

Sep 2025 - Ongoing

Prototyping and optimizing a real-time, multimodal AI navigation system integrating computer vision and audio processing. Constructing dynamic scene graphs alongside researchers to enable spatial reasoning for autonomous agents. This is in collaboration with KTH AI Society

Systems Simulation Engineer

AESIR (KTH Rocketry)

Sep 2025 - Ongoing

Developing the Freyja HIL (Hardware-in-the-Loop) system to simulate real-time aerospace sub-systems; performing rigorous Test Development & System Validation for embedded flight software to ensure mission reliability.

Summer Internship

Defence Research & Development Organisation (DRDO)

Jun 2024 - Jul 2024

Engineered an 8x60 phased array antenna and real-time GUI in MATLAB, achieving 25% efficiency gain and 15% loss reduction while performing radar degradation analysis for fault-tolerant system design.

Research Internship

TATA Memorial Hospital, ACTREC

Nov 2023 - Dec 2023

Deployed U-Net and YOLOv8L for multi-organ nuclei segmentation on the MoNuSeg dataset, achieving 30% higher accuracy with U-Net and 20% faster inference with YOLOv8L through performance trade-off analysis.

Research Internship

Indian Space Research Organisation (ISRO)

Sep 2023 - Oct 2023

Led the optimization of a 16x16 Phased Antenna Array for MEOSAR satellite signals, achieving >20% SNR improvement through Ansys HFSS simulations and beamforming algorithms (MVDR, LCMV).

Research Assistant

Indian Institute of Management (IIM)

Mar 2023 - Jun 2023

Developed Python scripts for ETL processes for 5,000+ audience responses and 200+ TV ads; applied NLP sentiment scoring and statistical modelling to quantify emotional impact on 10+ persuasion metrics.

// 04. DEPLOYED MODELS

AI Architectures

LungMate: AI-Driven Lung Cancer Diagnosis

Hybrid Computer Vision system for multi-metric tumor quantification, integrated with a PyQt GUI for real-time image enhancement and automated diagnostic visualization.

Hybrid Vision
EfficientNet-B3 Swin Transformer CBAM YOLOv11 PyQt

AI-Driven Protocol Diagnostic Suite

Architecting a Generative AI pipeline to automate the interpretation of LoRaWan/Physical layer anomalies in telemetry logs (Unstructured Data Processing) through RAG-based diagnostics.

RAG Pipeline
LangChain MLflow Generative AI
Patent Filed

Industrial IoT Safety Wearable

Engineered a real-time Embedded System on ESP32 utilizing Sensor Fusion (MQ2 Gas, NEO-6M GPS, Pulse Sensor) and LoRaWAN connectivity for critical safety telemetry in hazardous environments.

LoRaWAN
ESP32 Sensor Fusion C++
Patent Filed

High-Altitude Autonomous Safety System

Designed a real-time Embedded System for fall detection utilizing Sensor Fusion and automated actuator deployment for high-reliability safety applications.

Automated Actuation
Embedded Systems Hardware-in-the-Loop
// 05. CAPABILITIES

TOOLS & TECHNOLOGIES

Python
MATLAB
SQL
PyTorch
YOLO (v8/v11)
U-Net
LangChain (RAG)
Hugging Face
Computer Vision
GCP
MLflow
Git/GitHub
Linux
Agile
System Validation
Power BI
Tableau
LaTeX
// 07. R&D OUTPUT

Patents & Publications

Intellectual Property

PATENT FILED

IoT Helmet & Safety Wearable

Real-time Embedded System on ESP32 utilizing Sensor Fusion (MQ2 Gas, NEO-6M GPS, Pulse Sensor) and LoRaWAN connectivity.

PATENT GRANTED

High-Altitude Fall Protection

Real-time Embedded System for fall detection utilizing Sensor Fusion and automated actuator deployment for high-reliability safety applications.

Academic Literature

15+ Published Research Papers

Journals: IEEE, Springer, Elsevier

Encompassing advanced signal processing, medical computer vision, array optimization, and neural network architectures.

Recognitions

2× Raman Award Winner Excellence in research and innovation.
LLM Evaluation Participant Google DeepMind & Andon Labs (Paid Honorarium).
// 09. HANDSHAKE PROTOCOL

Initiate Contact

Let's build the future of AI.

Currently open to research collaborations, MLOps architect roles, and challenging engineering problems.