Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am an AI Engineer specializing in building and deploying intelligent systems that solve real-world challenges. With a Master’s from the University of Maryland, my work is centered on the cutting edge of Agentic AI, Deep Learning, and Computer Vision. I have a proven track record of designing high-impact models , from custom transformers for 3D object detection [cite: 51] and fine-tuned LLMs for explaining autonomous vehicle behavior , to advanced multimodal systems that fuse LiDAR and image data for enhanced perception. My experience extends to developing multi-agent systems using LangGraph and engineering custom data ingestion pipelines for client-specific knowledge bases. I am proficient in the entire MLOps lifecycle , leveraging tools like PyTorch, TensorFlow, and a range of AWS services to take complex data and transform it into innovative, production-ready solutions.…I am an AI Engineer specializing in building and deploying intelligent systems that solve real-world challenges. With a Master’s from the University of Maryland, my work is centered on the cutting edge of Agentic AI, Deep Learning, and Computer Vision. I have a proven track record of designing high-impact models , from custom transformers for 3D object detection [cite: 51] and fine-tuned LLMs for explaining autonomous vehicle behavior , to advanced multimodal systems that fuse LiDAR and image data for enhanced perception. My experience extends to developing multi-agent systems using LangGraph and engineering custom data ingestion pipelines for client-specific knowledge bases. I am proficient in the entire MLOps lifecycle , leveraging tools like PyTorch, TensorFlow, and a range of AWS services to take complex data and transform it into innovative, production-ready solutions.

Abubakar Siddiq





I am an AI Engineer specializing in building and deploying intelligent systems that solve real-world challenges. With a Master’s from the University of Maryland, my work is centered on the cutting edge of Agentic AI, Deep Learning, and Computer Vision. I have a proven track record of designing high-impact models , from custom transformers for 3D object detection [cite: 51] and fine-tuned LLMs for explaining autonomous vehicle behavior , to advanced multimodal systems that fuse LiDAR and image data for enhanced perception. My experience extends to developing multi-agent systems using LangGraph and engineering custom data ingestion pipelines for client-specific knowledge bases. I am proficient in the entire MLOps lifecycle , leveraging tools like PyTorch, TensorFlow, and a range of AWS services to take complex data and transform it into innovative, production-ready solutions.…I am an AI Engineer specializing in building and deploying intelligent systems that solve real-world challenges. With a Master’s from the University of Maryland, my work is centered on the cutting edge of Agentic AI, Deep Learning, and Computer Vision. I have a proven track record of designing high-impact models , from custom transformers for 3D object detection [cite: 51] and fine-tuned LLMs for explaining autonomous vehicle behavior , to advanced multimodal systems that fuse LiDAR and image data for enhanced perception. My experience extends to developing multi-agent systems using LangGraph and engineering custom data ingestion pipelines for client-specific knowledge bases. I am proficient in the entire MLOps lifecycle , leveraging tools like PyTorch, TensorFlow, and a range of AWS services to take complex data and transform it into innovative, production-ready solutions.

Available to hire

I am an AI Engineer specializing in building and deploying intelligent systems that solve real-world challenges. With a Master’s from the University of Maryland, my work is centered on the cutting edge of Agentic AI, Deep Learning, and Computer Vision.

I have a proven track record of designing high-impact models , from custom transformers for 3D object detection [cite: 51] and fine-tuned LLMs for explaining autonomous vehicle behavior , to advanced multimodal systems that fuse LiDAR and image data for enhanced perception. My experience extends to developing multi-agent systems using LangGraph and engineering custom data ingestion pipelines for client-specific knowledge bases.

I am proficient in the entire MLOps lifecycle , leveraging tools like PyTorch, TensorFlow, and a range of AWS services to take complex data and transform it into innovative, production-ready solutions.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Language

English

Fluent

Work Experience

AI Engineer at Easybee AI

August 1, 2024 - Present

Developing multi-agent systems using LangGraph and CrewAI by assigning specialized roles to agents to automate complex, interconnected tasks and enhance operational efficiency. Deployed the system to AWS ECS for automatic scaling. Engineered a proprietary custom web crawler and data ingestion pipeline to construct client-specific Knowledge Bases, implementing advanced chunking and data enhancement techniques before indexing into a Pinecone vector database for agentic AI access. Integrated and managed WhatsApp and Twilio messaging channels using Model Context Protocol (MCP) to enable seamless and continuous communication between customers and the AI system across different platforms.

Deep Learning Engineer at UMD PAL

May 1, 2025 - October 2, 2025

Developed a custom vision model based on Masked Autoencoders (MAE) trained on RGB orthomosaics from drones to estimate Leaf Area Index (LAI); achieved an R2 score of 0.87 using only RGB data. Designed a ViT-based multi-spectral pipeline that takes DSM & DTM as inputs and trained for plant height estimation, achieving an RMSE of 0.26 m. Automated large-scale aerial data preprocessing and normalization pipelines, integrating OpenCV, rasterio, and PyTorch for scalable experimentation.

Computer Vision Intern at Vyorius

November 1, 2021 - October 2, 2025

Built YOLO-based video analytics modules for UAVs, enabling real-time object detection, tracking, and classification from aerial streams. Used SOTA segmentation models and integrated them into Vyorius’s DataSync Intelligence pipeline, enhancing vision-based landing and contributing to faster anomaly detection and higher precision in GPS-denied landings. Partnered with engineering to integrate and deploy these computer vision modules onto the Vyorius cloud platform, enabling scalable, AI-driven analytics for multi-robot missions.

AI Engineer at Easybee AI

August 1, 2024 - Present

Developing multi-agent systems using LangGraph and CrewAI by assigning specialized roles to agents to automate complex, interconnected tasks and enhance operational efficiency. Deployed the system to AWS ECS for automatic scaling. Engineered a proprietary web crawler and data ingestion pipeline to construct client-specific Knowledge Bases, performing advanced chunking and data enhancement before indexing into a Pinecone vector database for agentic AI access. Integrated and managed WhatsApp and Twilio messaging channels using Model Context Protocol (MCP) to enable seamless cross-platform communication between customers and the AI system.

Deep Learning Engineer at UMD PAL

May 1, 2025 - October 2, 2025

Developed a custom vision model based on Masked Autoencoders (MAE) trained on RGB drone orthomosaics to estimate Leaf Area Index (LAI), achieving an R2 score of 0.87 using only RGB data, reducing reliance on multispectral hardware. Designed a ViT-based multispectral pipeline that takes DSM/DTM inputs for plant height estimation, achieving an RMSE of 0.26 m and outperforming CNN baselines. Automated large-scale aerial data preprocessing and normalization pipelines with OpenCV, rasterio, and PyTorch for scalable experimentation.

Computer Vision Intern at Vyorius

November 1, 2021 - October 2, 2025

Built YOLO-based video analytics modules for UAVs to enable real-time object detection, tracking, and classification from aerial streams, expanding autonomous surveillance coverage by 60% across remote deployments. Integrated state-of-the-art segmentation models into Vyorius’s DataSync Intelligence pipeline, improving vision-based landing and 25–40% higher precision in GPS-denied landings. Collaborated with engineering to deploy CV modules onto Vyorius cloud platform, enabling scalable, AI-driven analytics for multi-robot missions.