Skip to content

Role profile

AI Engineer

AI Engineer delivering enterprise-grade Generative AI for UAE energy, banking and government sectors. 4+ years building agentic systems, RAG pipelines and event-driven architectures, with direct delivery to Fertiglobe, RTA, TAQA, First Abu Dhabi Bank and Dubai Islamic Bank.

Get in touch

Agentic AI, RAG and LLMOps, built for production.

Revenue impact
USD 50M+
RAG latency cut, 15s to 4s
73%
Engineering hours saved a year
1000+
Connectors on NovaFloat
1500+
DB tables via text-to-SQL
500+
Legal cases indexed
300K+
Enterprise eval accuracy
95–98%
Cross-department productivity
65%+

Capabilities

What I deliver.

  • 01

    Agentic AI & multi-agent systems

    Orchestrator agents that delegate to scoped sub-agents on LangGraph, CrewAI and A2A, with custom MCP tool servers and tool-level guardrails.

  • 02

    RAG, text-to-SQL & knowledge graphs

    Multilingual EN/AR retrieval, hybrid search on Azure AI Search, Neo4j knowledge graphs and text-to-SQL across 500+ enterprise tables.

  • 03

    Azure AI Foundry & Copilot Studio

    Foundry agents with JSON-schema outputs and dedicated evaluation agents, plus grounded Copilot Studio agents deployed to Microsoft Teams.

  • 04

    MLOps, LLMOps & observability

    CI/CD and model versioning on Azure DevOps, deployment to Container Apps, Functions and AKS with Helm, and Langfuse tracing and evaluation.

  • 05

    Data engineering on Microsoft Fabric

    Lakehouse ETL into Delta tables, warehouse pipelines and Power BI reporting driven by Azure AI Foundry agents.

  • 06

    Fine-tuning & computer vision

    LoRA / QLoRA fine-tuning of SLMs, BERT classification, and YOLO + OCR document pipelines for bank-grade cheque processing.

Skill set

The toolkit behind the work.

Generative AI

13
  • Large Language Models
  • Agentic AI
  • Multi-Agent Orchestration
  • Sub-Agents & Skills
  • Harness Engineering
  • Context Engineering
  • Prompt Engineering
  • RAG
  • Guardrails
  • Fine-Tuning
  • LoRA / QLoRA
  • Embeddings
  • Semantic Search

LLM stack

9
  • LangChain
  • LangGraph
  • CrewAI
  • GPT-4o
  • Azure OpenAI
  • Claude
  • LLaMA
  • Hugging Face Transformers
  • BERT

Vector search & knowledge

8
  • Azure AI Search
  • Pinecone
  • FAISS
  • Chroma
  • Weaviate
  • Neo4j
  • Knowledge Graphs
  • Hybrid Search

Cloud, infra & MLOps

14
  • Microsoft Azure
  • Azure AI Foundry
  • Azure Machine Learning
  • Copilot Studio
  • Container Apps
  • Azure Functions
  • Cosmos DB
  • AKS
  • Helm
  • Docker
  • AWS
  • Azure DevOps
  • CI/CD
  • MLflow

Data engineering

9
  • Microsoft Fabric
  • ETL / ELT
  • Delta Tables
  • Azure Data Factory
  • Lakehouse
  • Data Warehouse
  • SQL
  • Text-to-SQL
  • Power BI

Protocols & governance

10
  • MCP
  • A2A
  • REST APIs
  • GraphQL
  • OAuth 2.0
  • Microsoft Entra ID
  • RBAC
  • Evaluation & Grounding
  • Responsible AI
  • Langfuse

Backend, ML & vision

10
  • Python
  • FastAPI
  • PyTorch
  • TensorFlow
  • NLP
  • PostgreSQL
  • Redis
  • YOLO
  • OCR
  • Signature Matching
See every skill with proficiency and years

Selected work

Case studies for this role.

View all projects

Results

Key achievements.

  1. 01Delivered 25+ production agentic AI systems across legal, banking, energy and government domains for UAE/GCC enterprise clients.
  2. 02Drove USD 50M+ revenue impact and saved 1000+ engineering hours annually through the Data2Decisions multi-agent platform.
  3. 03Achieved 95% extraction and 98.75% evaluation accuracy on the Fertiglobe Microsoft Fabric extraction pipeline.
  4. 04Cut multilingual RAG latency by 73% (15s to 4s) for 1000+ government users on RTA Mahboob.
  5. 05Pioneered production MCP and A2A adoption with tool- and agent-level guardrails.
Also seeAI Solution Architect & AI ConsultantGoverned, cloud-native AI on Azure, AWS and Google Cloud.