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

I'm Shamemdeen Allauteen, a senior AI and GenAI engineer based in Dubai. I design and deploy production-grade systems that blend LLMs with multimodal components, retrieval-augmented workflows, and real-time generative outputs. My focus is solving real-world business problems by building scalable, cost-efficient AI pipelines on Azure and GCP, leveraging advanced optimization techniques such as FlashAttention, quantization, and dynamic batching to reduce operating costs and accelerate research-to-production cycles. Over the past several years I’ve led end-to-end AI programs—from data ingestion and intent understanding to deploying high-throughput, observable AI services. I enjoy bridging cutting-edge research with robust MLOps, building hybrid RAG architectures, speech-to-speech and vision-language solutions, and delivering measurable impact across manufacturing, retail, and real estate domains.…I'm Shamemdeen Allauteen, a senior AI and GenAI engineer based in Dubai. I design and deploy production-grade systems that blend LLMs with multimodal components, retrieval-augmented workflows, and real-time generative outputs. My focus is solving real-world business problems by building scalable, cost-efficient AI pipelines on Azure and GCP, leveraging advanced optimization techniques such as FlashAttention, quantization, and dynamic batching to reduce operating costs and accelerate research-to-production cycles. Over the past several years I’ve led end-to-end AI programs—from data ingestion and intent understanding to deploying high-throughput, observable AI services. I enjoy bridging cutting-edge research with robust MLOps, building hybrid RAG architectures, speech-to-speech and vision-language solutions, and delivering measurable impact across manufacturing, retail, and real estate domains.

Shamemdeen Allauteen

AI Engineer, Data Scientist, Full Stack Developer, +2





I'm Shamemdeen Allauteen, a senior AI and GenAI engineer based in Dubai. I design and deploy production-grade systems that blend LLMs with multimodal components, retrieval-augmented workflows, and real-time generative outputs. My focus is solving real-world business problems by building scalable, cost-efficient AI pipelines on Azure and GCP, leveraging advanced optimization techniques such as FlashAttention, quantization, and dynamic batching to reduce operating costs and accelerate research-to-production cycles. Over the past several years I’ve led end-to-end AI programs—from data ingestion and intent understanding to deploying high-throughput, observable AI services. I enjoy bridging cutting-edge research with robust MLOps, building hybrid RAG architectures, speech-to-speech and vision-language solutions, and delivering measurable impact across manufacturing, retail, and real estate domains.…I'm Shamemdeen Allauteen, a senior AI and GenAI engineer based in Dubai. I design and deploy production-grade systems that blend LLMs with multimodal components, retrieval-augmented workflows, and real-time generative outputs. My focus is solving real-world business problems by building scalable, cost-efficient AI pipelines on Azure and GCP, leveraging advanced optimization techniques such as FlashAttention, quantization, and dynamic batching to reduce operating costs and accelerate research-to-production cycles. Over the past several years I’ve led end-to-end AI programs—from data ingestion and intent understanding to deploying high-throughput, observable AI services. I enjoy bridging cutting-edge research with robust MLOps, building hybrid RAG architectures, speech-to-speech and vision-language solutions, and delivering measurable impact across manufacturing, retail, and real estate domains.

Available to hire

I’m Shamemdeen Allauteen, a senior AI and GenAI engineer based in Dubai. I design and deploy production-grade systems that blend LLMs with multimodal components, retrieval-augmented workflows, and real-time generative outputs. My focus is solving real-world business problems by building scalable, cost-efficient AI pipelines on Azure and GCP, leveraging advanced optimization techniques such as FlashAttention, quantization, and dynamic batching to reduce operating costs and accelerate research-to-production cycles.

Over the past several years I’ve led end-to-end AI programs—from data ingestion and intent understanding to deploying high-throughput, observable AI services. I enjoy bridging cutting-edge research with robust MLOps, building hybrid RAG architectures, speech-to-speech and vision-language solutions, and delivering measurable impact across manufacturing, retail, and real estate domains.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Language

English

Fluent

Tamil

Fluent

Malayalam

Fluent

Hindi

Fluent

Arabic

Beginner

French

Beginner

Work Experience

AI & ML Engineer at RoyalClass

October 1, 2025 - Present

Architected and deployed a zero-shot plus fine-tuned multimodal content moderation system using open-source vision-language models (LLaVA-Next, Florence-2, CLIP-ViT) with custom LoRA adapters. Achieved 98.7% accuracy in detecting off-category and policy-violating listings. Engineered a hybrid rejection pipeline with rule-based filters, semantic similarity search (Sentence Transformers plus Chroma DB), and GPU-accelerated inference (TensorRT-LLM plus vLLM), reducing false positives by 42% and processing 10K+ daily listings under 800ms latency on A100 GPUs. Developed an end-to-end AI imaging suite with latent diffusion (Stable Diffusion XL inpainting/outpainting) and ControlNet-based enhancement models for automatic property photo enhancement and virtual staging, cutting manual retouching workload by 80%. Implemented a real-time AI voice assistant (direct speech-to-speech, no intermediate ASR) using TorToiSe plus CoquiTTS plus custom voice cloning, achieving <800ms end-to-end latency for

Generative AI & Machine Learning Engineer at Vienna Solutions

May 1, 2024 - September 1, 2025

Led development of real-time AI solutions for vending machine operations and AI procurement workflows. Ingested procurement data via Azure Data Factory, synchronized data across MongoDB and SQL stores with validation and schema checks. Optimized MongoDB data layer with adaptive indexing and caching, reducing average query latency by ~30% across 100K+ records. Implemented NLP-based query understanding using spaCy, Sentence Transformers, and LLM-assisted parsing via LangChain plus Azure OpenAI, achieving ~90-95% intent detection. Implemented a hybrid RAG retrieval pipeline using Azure AI Search (vector plus keyword search) combined with TF-IDF and fuzzy matching, improving recall by ~20% on complex multi-item queries. Built a supplier ranking service using the TOPSIS multi-criteria decision algorithm, improving alignment with expert selections by ~25%. Deployed a predictive maintenance system for 2,000+ IoT vending machines using Prophet and LSTM. Trained and deployed models using Azure

AI Engineer at Bright/River

November 1, 2022 - April 1, 2024

Led AI-driven image editing solutions for high-volume e-commerce content production. Built an AI agent system processing 10M+ product images daily using Stable Diffusion inpainting with 98.5% accuracy and CLIP-based classifiers, reducing manual review by 70%. Developed a high-performance CUDA and TensorRT inference pipeline for 1.2TB/day user content, integrating Vision Transformers (ViT), BERT+CRF (94% recall), and StyleGAN3 for adversarial robustness. Quantized models with TensorFlow Lite and ONNX Runtime for edge deployment, cutting cloud costs by 60% with sub-8ms latency on NVIDIA T4 GPUs. Built RAG agents trained on 50K manuals using LangChain and LlamaIndex to auto-resolve 40% of pipeline errors, reducing downtime by 35%. Led a seven-member team to productionize ML models with Azure ML, accelerating A/B testing velocity by 60% using TFX-based feature stores.

Data Engineer at Bright/River

February 1, 2021 - November 1, 2022

Streamlined cloud-based imaging automation for scalable e-commerce solutions. Engineered imaging automation platforms enabling 40% faster data retrieval through optimized pagination and asynchronous queries. Enhanced data extraction workflows, increasing customer satisfaction by 25% and reducing data quality issues by 30%. Managed imaging datasets with Azure Data Factory, incorporating advanced pattern recognition logic. Optimized batch image data processing for high-throughput pipelines and implemented robust data caching.

Software Engineer at Snapminds

January 1, 2020 - January 1, 2021

Developed AI and cloud-based solutions for complex recruitment processes; served clients including TVS, Puma, PepsiCo, and Accenture. Built real-time AI-powered recruitment platforms, optimized cloud-based workflows with Azure and AWS, and implemented automated data pipelines to reduce manual processing time by 30%. Delivered real-time analytics dashboards for recruitment metrics enabling data-driven decisions and reducing cycle time by 25%.