Building AI systems, agents, and workflows designed for real operations.

👨‍💻 AI systems engineer focused on deployable AI workflows, automation, and production-ready infrastructure for businesses that still run on messy processes and operational bottlenecks.

⚡ Currently working as a Junior AI Developer at Webdura Technologies, building AI-backed marketing systems, automation workflows, and agent assistants for traditional industries.

I work at the seam between:

• 🤖 AI systems
• ⚙️ workflow automation
• 🧠 product thinking
• ☁️ infrastructure
• 📈 operational scalability

My focus is not just building models — but building systems that survive production:

• 🌍 multilingual inputs
• ⚡ latency constraints
• 🚢 deployment realities
• 🔄 workflow integration
• 📊 operational scale

I’ve worked across:

• 🗣️ NLP
• 🤖 AI agents
• 🔎 retrieval systems
• 📈 forecasting
• 👁️ computer vision
• 📊 business intelligence
• ☁️ cloud-native AI deployments

…and I care deeply about AI that actually fits into how businesses operate.

• 🤖 AI agents & orchestration
• 🔎 Retrieval-Augmented Generation (RAG)
• ⚡ AI workflow automation
• ☁️ Production AI infrastructure
• 🛠️ MLOps & cloud deployment
• 🏢 Business-facing AI systems
• 🚀 Rapid AI prototyping & experimentation

📅 2026 → Present

Building AI-backed marketing products and AI agent assistants for traditional businesses.

• ⚙️ Designing AI-assisted operational workflows
• 🧩 Working closely with product ideation and business pain points
• 📈 Building scalable automation systems around AI primitives
• 🚀 Rapid R&D on deployable AI workflows

📅 2025

Built AI workflows and scalable cloud-native systems for recruitment operations across Gulf & Middle East markets.

• 🧠 Generative AI workflows
• 🔐 Authentication-integrated AI systems
• ⚡ Rapid AI prototyping pipelines

📅 2025

Owned NLP moderation infrastructure deployed on AWS.

• 🌍 Multilingual hate-speech detection
• ⚡ Real-time moderation pipelines
• 🐳 Dockerized inference systems
• ☁️ Production-scale deployment workflows

📅 2025

Worked on computer vision systems across:

• 🚨 theft detection
• 🚦 traffic analysis
• 🛰️ satellite imagery

Built few-shot learning pipelines using Vision Transformer backbones.

📅 2024 → 2025

Led AI-assisted content optimization workflows.

• 📈 Scaled organic impressions from 2.43K → 477K
• 🧠 Built analytics-driven operational processes
• 🤝 Managed AI-assisted content systems

Production-ready moderation system for multilingual toxicity and hate-speech detection.

• 🔥 PyTorch
• 🐳 Docker
• ☁️ AWS
• 📚 NLTK
• 🧠 Ensemble NLP models

• ⚡ Real-time inference pipeline
• 🌍 Multilingual handling
• 🛡️ Fault-tolerant architecture
• 🏠 Self-hosted inference stack
• 📝 Retraining-ready logging workflows
• 🔒 Privacy-first deployment approach

Built a business intelligence assistant allowing retail operators to query warehouse data in natural language.

• 🧠 Mistral-7B
• 🌶️ Flask
• 🗄️ SQL
• ☁️ AWS

• ⚡ Low-latency business querying
• 🔎 RAG-assisted retrieval
• 📈 Analyst-free operational insights

Decision-support forecasting system for retail operations.

• 🧮 TensorFlow
• 📈 LSTM
• 🎛️ Streamlit
• 📚 Scikit-learn

• 📊 Seasonality-aware forecasting
• 📈 Interactive dashboarding
• 🧪 Backtesting workflows

Generalized computer vision pipeline for low-data domains.

• 🔥 PyTorch
• 👁️ Vision Transformers
• 🎯 Few-shot learning

• 🛰️ Satellite imagery
• 🏪 Retail vision systems
• ⚡ Low-label adaptation workflows

I’m interested in AI systems that:

• ⚡ survive production
• 🏢 fit operational realities
• 🤖 automate real workflows
• 🔄 reduce friction inside businesses
• 📈 create leverage through infrastructure and intelligence

Benchmarks are useful.

Operational reliability matters more.

• 🚴 Cycling clears my head better than debugging
• 🎮 Strategy games are my favorite way to think through systems
• 📚 Usually reading AI papers, infra blogs, or startup/operator essays
• ☕ Most ideas start from overthinking workflows that could be automated

👨‍💻 Engineer by role. Systems thinker by instinct.