in developing efficient Real-time Computer Vision/Image processing/ Natural language processing algorithms using Deep learning/Traditional Machine learning
Robust acumen
Contributed to L4-Autonomous driving under percept…
in developing efficient Real-time Computer Vision/Image processing/ Natural language processing algorithms using Deep learning/Traditional Machine learning
Robust acumen
Contributed to L4-Autonomous driving under perception module covering camera/LiDAR based Semantic-Segmentation and 2D/3D Object-Detection, LiDAR based NDT Mapping and Localization
Contributed to Advanced Driver Assistance systems-ADAS (Driver Drowsy Detection, Day and Night Time Pedestrian Detection, Traffic Sign Detection, Object Tracking, Stereo Vision for Depth, Disparity & Object Detection) along with Machine Vision for mining industry related projects
Experience working with Structure-From-Motion, Bundle Adjustment, Multi-View-Stereo, Camera Relocalization, LiDAR-SLAM, RADAR-SLAM, Loop-closure, Camera-LiDAR calibration, LiDAR-IMU calibration, SLAM-backend optimization using GTSAM, Ceres
Experience of Text/Label Detection, Natural-language inference, Topic-modeling, Text Summarization, Named Entity Recognition, Question-Answering, Build & Fine-Tune State of the art LLM models on custom PDF/receipt/ticket forms, OCR Development
Experience working with Supervised-finetuning (SFT), Reward modeling (RM) and RLHF using open-source LLMs (LLama/ Mistral) on open-source datasets
Experience working with RAG technologies and LLM frameworks (Langchain and LLamaIndex), LLM model registries (Hugging Face), LLM APIs, LLM evaluation (perplexity, ROGUE, BLEU, BERTSCORE) embedding models, and vector databases (FAISS)
Skilled in applying methods to structured/unstructured problems using CNN, RNN, Geometric deep-learning, Transformers, Dimensionality reduction (PCA/t-SNE/UMAP) and clustering techniques
Knowledge of Few-shot/Zero-shot/Meta-learning/Lifelong-learning/Continual-learning/Out-of-distribution/camera-ISP pipeline for business use-case
Demonstrated advanced foundation in Linear Algebra, Calculus and Statistics
Knowledge of DNN performance optimization using ONNXRunTime, TensorRT (Fp32/Fixed-point/INT8 precision) covering: Quantization Aware Tuning; Post-Training Quantization (PTQ); Network Pruning; Machine learning compilers (TVM/GLOW);
Knowledge of Vectorization, Tiling, Parallelizing, Loop unrolling, CUDA-programming, Multi-threading/ Multi-processing/ Asynchronous programming
Worked on complex and rich data sets to build and deploy the best data solutions to complex business problems using cutting-edge machine learning, deep learning, and computer vision techniques to invent solutions
Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module lead Module leadModule lead Module lead Module lead Module lead
Worked on 2D object detection for warehouse customer
Gathered external datasets
Built synthetic datasets as per the business needs
Worked on Explainable AI
Worked on video reconstruction from event cameras using supervised/self-supervised/GAN and spiky-networks
Worked on Transformers
Worked on Knowledge distillation
Worked on XAI
Worked on HDR imaging
Worked on Graph networks
Worked on Structure-from-motion
Worked on Multi-view-stereo
Worked on Tracking and Mapping
Worked on DNN optimization using post-training quantization
Worked on Quantization aware fine-tuning (QAT) using TensorRT
Worked on PyTorch
Worked on TVM Relay
Worked on TVM AutoTVM/Autoscheduler
Worked on ONNXRunTime
Worked on NLP text classification
Worked on text summarization
Worked on Question and Answering
Worked on Topic Modeling
Worked on Natural language inference
Worked on Time-series
Conducted out-of-calibration detection
Conducted dynamic calibration between camera-LiDAR sensors
Worked on video surveillance application on Intel Open-vino framework
Worked on L4 Autonomous Driving
Detected 3D objects using LiDAR point cloud on single-frame/multi-frame
Tracked 3D objects using LiDAR point cloud
Detected traffic light using tensorflow object detection API
Integrated Apollo/Autoware algorithm for 3D object detection and traffic light
Worked on simulated autonomous driving using Reinforcement Learning Framework with RayLib and tensorflow
Partook in competitions SemanticKITTI/Nuscenes/Argoverse/KITTI
Worked on LiDAR based mapping and localization for autonomous driving
Conducted DNN optimization using post-training quantization
Conducted Quantization aware fine-tuning (QAT) using TensorRT C++
Used PyTorch
Used ONNXRunTime
The AI-first Recruitment Operating System that unifies and automates the entire hiring lifecycle for modern enterprises worldwide.
Dr Job FZ LLC. 2026 © All Rights Reserved
