I'm an AI Engineer specializing in production-grade LLM systems, real-time voice AI, and inference optimization. I don't just prototype — I build and ship systems that run in the real world under real constraints.
- 🏛️ Built a government-grade real-time translation system handling 30+ languages with sub-1500ms latency (Morocco GovTech)
- 🎤 Designed scalable Voice AI platforms with inbound/outbound telephony via SIP routing (US project)
- ⚡ Currently deploying and optimizing LLMs at scale using vLLM, KV cache, and quantization for a London-based AI company
- 🤖 Experienced with multi-agent orchestration: LangGraph, CrewAI, AIBrix (TikTok's agent layer)
- 🔬 Fine-tuned LLMs and Vision-Language Models on proprietary datasets for domain-specific tasks
- 🌍 Remote AI Engineer working across UK, US, and international GovTech projects
"I care about AI that works in production — low latency, high reliability, real impact."
🏛️ Gov Real-Time Translation → 30+ languages · sub-1500ms · LiveKit + PipeCat + Whisper
🎤 Voice AI Platform (US) → SIP telephony · no-code agent builder · LangGraph agents
⚡ LLM Inference at Scale → vLLM · KV cache · FastAPI · LMCache via Redis
🔬 Vision-Language Fine-Tuning → Llama 3.2 Vision · Unsloth · medical/domain datasets
🤖 Multi-Agent Orchestration → AIBrix · CrewAI · LangGraph · tool-use agents
|
Fine-tuning pipeline using Unsloth for efficient LLM training with LoRA/QLoRA. Domain-specific fine-tuning with experiment tracking.
|
Fine-tuned Llama 3.2 Vision on radiography datasets for medical image analysis using Unsloth optimization.
|
|
Lightweight Text-to-Speech pipeline with multiple voice engine support and real-time audio processing.
|
AI-powered research assistant with document retrieval, synthesis, and structured output generation.
|
| Role | Company | Location | Period |
|---|---|---|---|
| 🤖 AI Engineer — LLM Systems | Confidential | London, UK 🇬🇧 | Jul 2025 – Present |
| 🌍 AI Voice Agent Developer | GovTech Project | Morocco 🇲🇦 | Feb – Jun 2025 |
| 🎤 AI Engineer — Voice AI Platform | US Project | Miami, USA 🇺🇸 | Nov 2024 – Feb 2025 |
| 🔬 Junior AI Engineer | Iaxon Software | Pakistan 🇵🇰 | Apr – Oct 2023 |
specializations = {
"Voice AI": ["PipeCat", "LiveKit", "Whisper ASR", "SIP Telephony", "TTS Pipelines"],
"LLM Inference": ["vLLM", "PagedAttention", "KV Cache", "GPTQ", "AWQ", "Quantization"],
"Fine-Tuning": ["LoRA", "QLoRA", "RLHF", "Unsloth", "SFTTrainer", "Vision-LMs"],
"Agentic Systems": ["LangGraph", "CrewAI", "AIBrix", "Tool Use", "Multi-Agent Orchestration"],
"RAG & Retrieval": ["LangChain", "Pinecone", "Weaviate", "RAGAS", "Hybrid Search"],
"Languages_spoken": ["Urdu 🇵🇰", "English 🇬🇧 (IELTS 7.5 / C2)"],
}I'm open to remote AI engineering roles and consulting contracts with US/UK/EU companies.
Specializing in: Voice AI systems · LLM deployment & optimization · Fine-tuning pipelines · Multi-agent architectures
"Building AI systems that work at government scale, voice latency, and production reliability."