ZEESHAN IBRAR inboxtozeeshanibrar@gmail.
com
+92 305 5057700
Portfolio: [Link] [Link]
GitHub: [Link]
PROFESSIONAL SUMMARY
Python AI/ML Developer with 2+ years building production-grade ML systems, RAG pipelines, and scalable backends. Skilled across the
full ML lifecycle — from data engineering and model training to LLM integration and deployment. Passionate about applying AI to real-
world products.
SKILLS
• Languages & Frameworks: Python, JavaScript, SQL, Django, Flask, FastAPI, React
• AI / ML: TensorFlow, Keras, PyTorch, Scikit-learn, NLTK, LlamaIndex, MLflow, RAGAS, RAG Pipelines, Prompt Engineering, Redis,
Vector Databases, Fine-tuning, LangChain, LLM APIs (OpenAI/Gemini)
• Data, Cloud & Tools: Pandas, PostgreSQL, MySQL, ETL Pipelines, AWS, Docker, Git/GitHub, IBM AS/400
WORK EXPERIENCE
EURONET PAKISTAN: Associate Software Developer July 2024 – Present
• Designed and maintained large-scale ETL and data reconciliation pipelines processing millions of financial records with zero-error
accuracy across production cutover cycles
• Engineered automated data migration utilities for bank onboarding, reducing manual effort and enabling repeatable, auditable
workflows across diverse customer datasets
• Built multiprocessing-based automation tools for high-volume data operations, improving throughput and reliability in time-
sensitive financial processing environments
NIXAAM LLC: Mid-Level Python AI/ML Developer January 2026 - April 2026 (Part-time)
• Built and deployed applied ML solutions for classification, prediction, and NLP-based tasks using PyTorch, TensorFlow, and Scikit-learn
in a production healthcare environment
• Built SnipTool, a production Windows desktop app using PySide6 integrating live-annotated screen recording, [Link] offline
speech-to-text, and GPT-4o-mini for automated meeting notes generation with mixed Urdu/English support
PROJECTS
GLAMORA (AI FASHION RECOMMENDER ASSISTANT) [Link]
• Solved the cold-start recommendation problem by profiling new users via a style quiz, eliminating the need for purchase history to
generate relevant suggestions
• Built a hybrid ML pipeline (TF-IDF → k-means clustering → cosine similarity) achieving 89% recommendation relevance across
1,000+ user profiles
• Delivered full-stack via Django REST + React, with Razorpay payment integration and mobile-responsive UI
RAGSYNAPSE (PRODUCTION DOCUMENT INTELLIGENCE) [Link]
• Built a production-grade Multi-LLM RAG system using LlamaIndex, FastAPI, and Redis, achieving 0.87 RAGAS faithfulness score
with full MLflow observability and 5-service Docker Compose deployment
• Engineered automatic rate-limit fallback and per-provider cost tracking across multiple LLM providers (OpenAI/Gemini), enabling
reliable and cost-aware document intelligence at scale
• Forked and optimized an open-source RAG base into a production-ready system, delivering precise page-level citations from any
document with end-to-end traceability and measurable RAG performance
EDUCATION HONORS AND AWARDS
2ND BOARD POSITION IN HSSC, SINDH
BSE, COMPUTER SOFTWARE ENGINEERING Bahria College Karsaz • 2019–2021
Bahria University • 2021–2025
CERTIFICATIONS
MACHINE LEARNING A-Z: AI, PYTHON & R + CHATGPT PRIZE [2024] Udemy • 2024
100 DAYS OF CODE: THE COMPLETE PYTHON PRO BOOTCAMP Udemy • 2023