AVAILABLE · AI ENGINEER ROLES

ANISHAN
SEBANATHAN

AI ENGINEER/ML SPECIALIST

GT M.S. AEROSPACE · ENSEEIHT M.S. CS · GPA 4.0

VIEW PROJECTSCONTACT
4.0GPA / 4.0
2×M.S. DEGREES
4SYSTEMS BUILT
SCROLL
— DUAL ACADEMIC CORE —

DUAL‑CORE ADVANTAGE

GEORGIA TECHM.S. AEROSPACE ENG.
INP-ENSEEIHTM.S. COMPUTER SCIENCE
AI CORE

CS foundation from ENSEEIHT + Aerospace Engineering from Georgia Tech — algorithms, data systems, ML at scale, meeting autonomous systems and real-time perception.

MACHINE LEARNING
Predictive modeling · Large-scale data · End-to-end pipelines
PERCEPTION SYSTEMS
Real-time sensor fusion · SLAM · State estimation (EKF)
SYSTEMS THINKING
ATA22/27 validation · Embedded constraints · Hardware-aware AI
4.0
GPA / 4.0
5+
LANGUAGES
2×
M.S.
— PROFESSIONAL EXPERIENCE —

WHERE DATA MEETS IMPACT

INDUSTRY · MACHINE LEARNING

AIR FRANCE INDUSTRIES

Data Scientist & ML Engineer
MAY 2026 — DEC 2026
  • —Developed ML models for large-scale transportation cost prediction across logistics operations
  • —Designed end-to-end data pipelines processing operational data at enterprise scale
  • —Built interactive Streamlit dashboard for real-time data visualization and decision support
  • —Applied data-driven optimization to reduce operational costs during initial deployment
0M€
OPERATIONAL DATA PROCESSED
€0K
COST SAVINGS ACHIEVED
AEROSPACE · AUTOMATION

AIRBUS

Flight Systems Validation Intern
JUN 2024 — AUG 2024
  • —Automated compliance analysis for Airbus ATA22 (auto-flight) and ATA27 (flight controls) systems
  • —Developed Python tools for systematic test processing and traceability across validation workflows
  • —Improved reporting workflows by rationalizing data extraction and decision documentation
— SELECTED PROJECTS —

SYSTEMS BUILT, PROBLEMS SOLVED

Click any project to expand technical depth.

ML / DATA SYSTEMS

LARGE-SCALE COST PREDICTION SYSTEM

Air France Industries · Production · 2026

End-to-end ML system for transportation cost prediction across aviation logistics at enterprise data scale.

PREDICTIVE MLDATA PIPELINESTREAMLIT€87M SCALE
›Designed and trained supervised ML models (gradient boosting + regression ensembles) for multi-dimensional cost forecasting
›Built ETL pipelines to ingest, clean, and transform operational data from heterogeneous logistics sources
›Architected a feature engineering layer capturing temporal, route-level, and fleet-type signals
›Deployed an interactive Streamlit dashboard enabling finance and ops teams to explore predictions and run scenarios
STACKPython · Scikit-learn · Pandas · Streamlit · SQL
SIMULATION / SoS

WAARDEN — WILDFIRE DECISION SUPPORT SYSTEM

ASDL Grand Challenge · SAAB Technologies · Georgia Tech

Multi-agent System of Systems simulation for wildfire response optimization — coordinating drones, aircraft and ground teams under fog-of-war conditions, sponsored by SAAB Technologies.

MULTI-AGENT SIMSYSTEM OF SYSTEMSDOESTOCHASTIC
›Implemented a Python-based multi-agent simulation environment on the COLOSSUS SoS framework, coupling stochastic fire propagation (cellular automata, wind & terrain effects) with agent coordination
›Designed a centralized command architecture with agent–commander communication and decision-making under uncertainty, modeling real fog-of-war operational constraints
›Ran a Design of Experiments (~2 500 simulations) to characterize performance across agent configurations and environmental conditions
›Built an interactive dashboard for trade-off exploration across cost, efficiency, and burned-area metrics — delivering actionable decision support for fire operations
STACKPython · COLOSSUS · Agent-based Modeling · Stochastic Simulation · DOE · Data Visualization
PERCEPTION / SLAM

GPS-DENIED NAVIGATION SYSTEM

ASDL Grand Challenge · ONR-Funded · Georgia Tech

Real-time visual perception and localization pipeline for a marine autonomous vehicle operating in GPS-denied environments.

ORB-SLAM3ROS2PERCEPTION PIPELINEMPC CONTROL
›Implemented ORB-SLAM3 within a modular ROS2 architecture enabling real-time feature extraction, tracking, and map construction
›Built a multi-modal sensor fusion layer combining LiDAR point clouds and camera data for robust state estimation
›Integrated ArduPilot with a Model Predictive Control (MPC) layer for trajectory following under physical constraints
›Developed a full Gazebo simulation environment for sensor characterization and navigation stress-testing
STACKPython · C++ · ROS2 · Gazebo · OpenCV · ArduPilot
STATE ESTIMATION

AUTONOMOUS LOCALIZATION — EKF PIPELINE

Bubblerob · ROS2 · Georgia Tech

Extended Kalman Filter for real-time probabilistic state estimation and SLAM in a mobile robot system.

EKFSENSOR FUSIONSLAMSTATE ESTIMATION
›Implemented a full EKF pipeline fusing odometry, IMU, and LiDAR measurements into a consistent robot pose estimate
›Developed a Scan Matching-based SLAM frontend for incremental map building with loop closure detection
›Designed sensor fusion architecture handling asynchronous inputs at different frequencies with uncertainty propagation
›Built autonomous navigation behaviors (obstacle avoidance, goal pursuit) on top of the localization stack
STACKPython · ROS2 · NumPy · LiDAR · IMU
— TECHNICAL ARSENAL —

TOOLS & TECHNOLOGIES

MACHINE LEARNING
PythonScikit-learnTensorFlowPredictive ModelingSupervised LearningData AnalysisFeature EngineeringPandas / NumPy
PERCEPTION & ROBOTICS
ROS2ORB-SLAM3Extended Kalman FilterSensor FusionLiDAR / CameraOpenCVGazebo SimulationArduPilot
SOFTWARE ENGINEERING
C / C++JavaMATLABADAGitLinuxStreamlitSystem Integration
AI SYSTEMS (EXPLORING)
Generative AILLM-based SystemsMPC ControlStochastic ModelingBig Data PipelinesSQL
— LANGUAGES —
FRENCHNATIVE
ENGLISHC1
TAMILNATIVE
GERMANB1
JAPANESEA2/B1
— GET IN TOUCH —

LET'S BUILD

Open to AI Engineer, ML Engineer, and Research Engineer roles at forward-thinking teams. Available from late 2026.

EMAIL
anishan.sebanathan@gmail.com
GITHUB
github.com/Neosilver09
LINKEDIN
linkedin.com/in/anishan-sebanathan/
BASED IN ATLANTA, GA / TOULOUSE, FR·GEORGIA TECH · INP-ENSEEIHT·OPEN TO RELOCATION