Aditya Karan profile photo 1 on Tal

Aditya Karan

AI/ML Engineer | Data Analytics Consultant | SQL | Power BI | Machine Learning | Computer Vision | Ex-DRDO Research Fellow

Ahmedabad

< 1 year of work experience

Self-Employed
Dec 20252 mosJan 2026

➜ Engineered a real-time drone vs. bird classifier on live video using YOLOv8 + OpenCV; achieved high-precision detection by injecting physics-informed kinematic constraints (speed envelope, trajectory smoothness, oscillation frequency) into post-processing — going beyond vanilla bounding-box inference. ➜Reduced false positive rate significantly vs. baseline YOLO inference by filtering detections against physics-derived motion signatures, validated on real-world outdoor footage and a custom-annotated dataset. ➜Built full IoT alert layer: ESP32 microcontrollers trigger real-time Telegram API notifications upon threat classification — complete end-to-end pipeline from video ingestion → detection → classification → dispatch. ➜Architected the entire pipeline as a modular, deployable system: video ingestion → frame preprocessing → YOLOv8 inference → kinematic filter → classification → IoT dispatch. Validated on real-world footage.
DRDO, Ministry of Defence, Govt. of India
Jun 20243 mosAug 2024

Delhi, India

➜Designed an end-to-end ML pipeline for land-cover anomaly detection in satellite imagery targeting active border surveillance — one of a small number of undergraduate researchers selected for DRDO's classified research track. ➜Processed and analyzed large-scale multi-temporal remote sensing datasets using ArcGIS, QGIS, and Google Earth Engine; built feature extraction workflows to identify statistically significant spatial change patterns across time-series imagery. ➜Built spatial risk models integrating multi-source geospatial signals; delivered structured technical reports and documentation adopted directly into DRDO project workflows. ➜Co-authored peer-reviewed research published in IJERT (DOI: 10.5281/zenodo.18441874) — "Scenario-Driven Land Change Anomaly Assessment for Border Surveillance," 2025.
Jun 20241 moJun 2024

Rajasthan, India

Represented DRDO in cross-organizational technical sessions on environmental impact assessment and low-emission engineering strategies; contributed to collaborative working groups with multi-disciplinary teams.

Education

SILVER OAK UNIVERSITY

Bachelor of Technology - BTech, Computer Science

SILVER OAK UNIVERSITY

20224 yrs2026

Join Aditya and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Aditya Karan on Tal

Headline

AI/ML Engineer | Data Analytics Consultant | SQL | Power BI | Machine Learning | Computer Vision | Ex-DRDO Research Fellow

About

I am a Machine Learning Engineer and Data Analytics Consultant with experience in Computer Vision, Geospatial AI, SQL and Business Analytics. During my research fellowship at DRDO, I worked on geospatial anomaly detection, satellite imagery analysis and predictive modelling for real-world applications. I have also developed end-to-end Computer Vision and IoT-integrated systems using YOLOv8, OpenCV and Flask. My expertise includes: • Machine Learning • Data Analytics • SQL • Power BI • Predictive Analytics • Computer Vision • Flask Applications I am currently open to: • AI/ML Engineering Opportunities • Data Analytics Roles • Freelance Consulting Projects • Research & Development Collaborations Portfolio: https://refined-personal-brand.lovable.app/ Let's connect if you're looking for data-driven solutions or AI-powered systems.

Location

Ahmedabad

Work experience

Physics-Informed Real-Time Drone Detection System at Self-Employed, Dec 2025 - Jan 2026 · 2 mos, ➜ Engineered a real-time drone vs. bird classifier on live video using YOLOv8 + OpenCV; achieved high-precision detection by injecting physics-informed kinematic constraints (speed envelope, trajectory smoothness, oscillation frequency) into post-processing — going beyond vanilla bounding-box inference. ➜Reduced false positive rate significantly vs. baseline YOLO inference by filtering detections against physics-derived motion signatures, validated on real-world outdoor footage and a custom-annotated dataset. ➜Built full IoT alert layer: ESP32 microcontrollers trigger real-time Telegram API notifications upon threat classification — complete end-to-end pipeline from video ingestion → detection → classification → dispatch. ➜Architected the entire pipeline as a modular, deployable system: video ingestion → frame preprocessing → YOLOv8 inference → kinematic filter → classification → IoT dispatch. Validated on real-world footage.

Research Fellow — Geospatial AI & Remote Sensing at DRDO, Ministry of Defence, Govt. of India, Jun 2024 - Aug 2024 · 3 mos, Delhi, India, ➜Designed an end-to-end ML pipeline for land-cover anomaly detection in satellite imagery targeting active border surveillance — one of a small number of undergraduate researchers selected for DRDO's classified research track. ➜Processed and analyzed large-scale multi-temporal remote sensing datasets using ArcGIS, QGIS, and Google Earth Engine; built feature extraction workflows to identify statistically significant spatial change patterns across time-series imagery. ➜Built spatial risk models integrating multi-source geospatial signals; delivered structured technical reports and documentation adopted directly into DRDO project workflows. ➜Co-authored peer-reviewed research published in IJERT (DOI: 10.5281/zenodo.18441874) — "Scenario-Driven Land Change Anomaly Assessment for Border Surveillance," 2025.

DRDO Representative — Sustainable Development Technical Sessions at AAEIT, Jun 2024 - Jun 2024 · 1 mo, Rajasthan, India, Represented DRDO in cross-organizational technical sessions on environmental impact assessment and low-emission engineering strategies; contributed to collaborative working groups with multi-disciplinary teams.

Education

Bachelor of Technology - BTech, Computer Science, SILVER OAK UNIVERSITY