Amit B.

Machine Learning Engineer

Bengaluru, Karnataka, India13 yrs 7 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Expert in end-to-end ML pipeline development
Proficient in deep learning for computer vision and NLP
Innovative projects in Generative AI and action recognition

Stackforce AI infers this person is a Machine Learning Engineer specializing in Computer Vision and Generative AI.

Contact

Skills

Core Skills

Machine LearningDeep LearningGenerative AiData ScienceComputer VisionDigital Image Processing

Other Skills

Problem SolvingPyTorchGesture RecognitionMLOpsConvolutional Neural Networks (CNN)PythonDeepspeedRAGStable DiffusionObject DetectionSemantic segmentationLarge Language Models (LLM)Image SegmentationImage RegistrationTransformer Models

About

Software engineer skilled in datascience and end-to-end ML pipeline development. Proficient with ML frameworks like PyTorch, Tensorflow and deployment tools like DeepStream, TensorRT and Docker. Experienced in application of deeplearning for computer vision , NLP and GenAI applications. Some of the key projects include: - Adaptation of Time-Space Transformer (TimeSformer) for action recognition in soccer videos. - Development and successful deployment of hand pose and gesture recognition in driver assistance systems. Currently working on following Gen AI projects : - Designing efficient QnA systems that leverage Retrieval Augmented Generation (RAG) on Large Language Models (LLMs) - Implementing stable diffusion models for synthetic data generation. Passionate about applying machine learning to drive incremental improvements in the existing systems.

Experience

13 yrs 7 mos

Total Experience

4 yrs

Average Tenure

1 yr 5 mos

Current Experience

Intuit

Staff Software Engineer ( AI / ML )

Dec 2024 – Present · 1 yr 5 mos · Bengaluru, Karnataka, India · Hybrid

Machine LearningDeep Learning

Intel corporation

Staff Software Engineer ( Applied ML )

Mar 2024 – Dec 2024 · 9 mos · Bengaluru, Karnataka, India · Hybrid

Generative AIMachine Learning

Deep learning analytics

Senior ML Engineer ( independent contractor )

Sep 2018 – Dec 2023 · 5 yrs 3 mos · Remote

Developed a Timesformer ( temporal vision transformer) based model for soccer player action recognition.
Adapted the Detection Transformer (Detr) model to identify players and their respective teams.
Implemented a multi-camera player tracking algorithm.
Optimized the models using ONNX/Nvidia-tensorRT and orchestrated a Nvidia Deepstream pipeline for near real-time processing of multicamera 4K streams.
Oversaw the development, testing, deployment and regualar improvements of the end to end pipeline.
As a part of ongoing research, exploring stable diffusion for synthetic data generation aimed at reducing data collection cost and improving player detection.
Enhanced the enterprise LLM-RAG system with hybrid semantic search and conversation handling features.
Implemented an AI pipeline supporting functionalities like pedestrian and vehicle tracking, crowd density detection, speed detection and face masking. Utilized transfer learning on state-of-the-art models such as YOLOv3, MaskRCNN, and Crowdnet

Problem SolvingData ScienceMachine Learning

Mercedes-benz research and development india

2 roles

Senior ML Engineer

Promoted

Feb 2016 – Sep 2018 · 2 yrs 7 mos · On-site

Developed deeplearning based computer vision models for occupancy detection and hands localization.
Played an active role in training, testing and improving hand pose models for gesture recognition.
Demonstrated ownership in orchestrating the end-to-end process of continuous training, testing and delivery of multiple ML models for deployment. Mentored and led a team of 3 throughout the cycle.
Developed in-house tools for seamless data extraction, transformation, analysis and data mining.
Actively contributed to ML research by surveying the latest models, generating patent ideas to advance innovation in the field.

PyTorchConvolutional Neural Networks (CNN)Machine LearningComputer Vision

Engineer ( Digital Image Processing )

Jul 2012 – Feb 2016 · 3 yrs 7 mos · On-site

Applied digital image processing algorithms to perform structural analysis on microscopic images.
Implemented tailored algorithms for precise image registration and automatic segmentation.
Packaged the implemented algorithms in to a python-based tool for swift analysis of material images. This resulted in significant time and cost savings compared to traditional lab testing.
As an individual contributor, effectively oversaw development, release, communication, and collaboration with multiple international stakeholders.

PythonDigital Image Processing