Skip to main content
Abdullah Ramzan

Abdullah Ramzan

AI ENGINEER · LLM SYSTEMS · FULL-STACK ML DEPLOYMENT

Talent
LahoreMember since May 22, 2026
PythonFastAPITensorFlowKerasStreamlitOllamaVector EmbeddingsSemantic SearchMiniRocketRidgeClassifierCNNsSTFT Signal ProcessingDockerPostgreSQLRedisOpenAI APIOpenAI GPTComputer Vision
·

Hire this person

Send a job offer directly to this candidate

PythonFastAPITensorFlowKerasStreamlit

About

AI Engineer with 1+ year shipping production LLM and ML systems at OneClout. Designed, built, and deployed four end-to-end AI products — spanning real-time time-series classification, RAG pipelines, GPT-powered document parsing, and desktop automation — each replacing a manual workflow with a measurable business outcome. Seeking a remote AI Engineer role where ownership, speed, and full-stack delivery matter.

Core stack: Python · FastAPI · OpenAI API · RAG · MiniRocket · TensorFlow · Docker · PostgreSQL · Redis.

Experience

Associate AI/ML Engineer - OneClout

(2025-07)

  • Delivered Truckast — a real-time drum-state classifier (Waiting → Pouring → Washing) using STFT signal segmentation + MiniRocket + RidgeClassifier, achieving ~80% accuracy on unseen production data and eliminating a fully manual inspection process across live sensor streams.
  • Reduced false positives in operational dashboards by redesigning time-series labeling with a forward-labeling strategy, directly improving status-transition reliability for business-facing reports.
  • Shipped ChatRAG — a production RAG chat system achieving sub-100ms token streaming via FastAPI + WebSocket, Ollama LLM backend, and Redis-backed session memory with TTL management and semantic vector search.
  • Built and deployed a GPT-powered Fitness Chatbot on FastAPI with streamed responses, dynamic prompt construction, and contextual memory — live in a production backend serving personalized workout and diet plans.
  • Architected a modular FastAPI + PostgreSQL backend enabling plug-and-play LLM provider swaps with zero service downtime, future-proofing the platform against model churn.

Remote Machine Learning Intern - AI Community of Pakistan (AICP)

(2024-07 - 2024-09)

  • Completed 100% of task-based ML challenges on time in a fully async remote environment.
  • Applied feature engineering and anomaly detection on independent datasets, evaluated against AICP benchmarks.

Education

B.Sc. - Business & Information Technology - University of Engineering & Technology, Lahore (2020 - 2024)

Skills

AI ENGINEERLLM SYSTEMSFULL-STACK ML DEPLOYMENT

Reviews

Related