Zulqarnain Khan
Islamabad, Pakistan
+92 3177896472
zulqar4791@gmail.com
AI Engineer with 3+ years shipping production RAG pipelines, multi-agent systems, and LLM fine-tuning across FinTech, EdTech, and enterprise SaaS. I build LangChain/LangGraph orchestration, MCP agents, and vector search at scale—not slide-deck demos.
Technical Skills
Languages
PythonSQLC++
AI & LLM
LangChainLangGraphRAGMCPLoRA/QLoRARLHFDSPyLiteLLM
Backend
FastAPICeleryRedisKafkaDockerKubernetesGitHub Actions
Cloud & MLOps
AzureGCPAWSMLflowLangfuseLangSmithvLLMHugging Face
Soft Skills
Agile DeliveryCross-functional CollaborationProduction DebuggingSystem Design
Work Experience
AI Engineer
Wanclouds Inc
- Built multi-agent systems with MCP and LangGraph on enterprise knowledge bases.
- Maintained RAG pipelines (ChromaDB, FAISS, Azure AI Search) over 1,000+ documents.
- Deployed AI services on Kubernetes with zero-downtime CI/CD via GitHub Actions.
- FastAPI microservices with WebSockets; Langfuse for latency and token tracking.
Associate NLP Engineer
Developers Den LLC
- Multi-modal RAG for US mortgage PDFs—10,000+ documents, 500+ concurrent users.
- Reduced ingestion time 50% with parallel async workers; GraphRAG with Neo4j.
- Kafka real-time ingestion; LangSmith/Langfuse production monitoring.
AI Developer
Maktek.ai
- Fine-tuned LLaMA-3, Mistral, Gemma with LoRA/QLoRA; published on Hugging Face.
- vLLM serving with GPTQ/AWQ—3× throughput, ~50% lower hosting cost.
- RLHF alignment and synthetic data via model distillation (−60% labeling cost).
Education
Bachelor of Computer Software Engineering (Honours)
University of Engineering and Technology, Mardan
Mardan, Pakistan
Machine Learning Specialization
Stanford University (Coursera)
Online
Certifications
Languages
EnglishProfessional
UrduProfessional
PashtoNative
Interests
RAG SystemsAgentic AILLM Fine-TuningMLOpsFinTech AIEdTech