Industry Experience

Anyware Robotics

Senior Perception Engineer (Team Lead) [Jan 2026 – Present]

Rebuilt and grew the perception team to 5 engineers; own project scope and roadmap across 3D perception, camera systems, and learning-based work, with day-to-day technical mentorship.
Architected a custom dual-head, NMS-free segmentation model end-to-end, adding a wireframe pose-detection head for direct 3D box pose estimation — achieving 99.4 mAP@50 (+1.7 pp) and 98.7 mAP@50–95 (+2.4 pp) while cutting inference from 56 ms to 7 ms.
Overhauled the 2D/3D perception pipeline (parallelized 3-camera preprocessing, GPU-accelerated segmentation, refactored point-cloud stack), cutting 2D compute 900 ms → 90 ms and 3D 2 s → 300 ms; hid perception behind arm motion to bring the pick cycle from 12 s to 9 s and lift throughput by 100+ cases/hour.

Perception Engineer II [Jan 2024 – Dec 2025]

Shipped an end-to-end palletizing perception stack: three RGB-D cameras reconstruct the work cell in 0.1 s, enabling a FANUC CRX25 to palletize boxes in 6.3 s each at 573 boxes/hour.
Built an active-learning MLOps workflow with ensemble uncertainty scoring using foundation models for zero-shot labeling via Label Studio on GCP & Ray — reducing manual annotation by 40% and test-to-deploy cycles to 40 minutes.
Scaled training data via synthetic generation: Blender photogrammetry + StyleGAN3 domain randomization (21 styles) expanded the dataset 8× and improved detection mAP by +4.7 pp.
Deployed optimized models with ONNX + TensorRT on GPU edge devices for real-time inference, and orchestrated a ROS 2 Humble behavior-tree and HFSM control stack with real-time fault classifiers, reducing unplanned robot downtime by 35% across pilot cells.

Data Science Intern [July 2022 - August 2022]

Tech Stack: Catia, SolidWorks, Coppelia Sims (V-Rep Simulator), Python (PyBullet), C/C++.
Deployed a quality control automation system for glass defect detection.
Created Python scripts to generate intelligence reports from export databases.

Data Science Intern [January 2022 - June 2022]

Tech Stack: Python (PySpark, PyTorch), C/C++, Congos, PowerBI, AWS, Git, IBM-Cp4d.
Enhanced preprocessing pipeline with PySpark, improving speed and efficiency.
Built a database for electric vehicle charging station locations.
Developed a warning system for vehicle failure prediction.

Summer Intern [July 2021 - August 2021]