Hi, I'm Nadhif Faris A.S
AI/ML Engineer & Data Scientist
I build intelligent AI systems by combining machine learning, Large Language Models (LLMs), and data-driven solutions to solve real-world business challenges.
About Me
A quick look at my background and where I focus my work.
AI/ML Engineer and Data Scientist with hands-on experience spanning production ML systems, causal inference, and LLM-powered applications, backed by a Microsoft Azure AI Engineer Associate credential and enterprise data exposure through a Telkomsel internship. Skilled across the full delivery lifecycle, from data quality diagnosis and model design to RAG pipelines, agentic workflows, and cloud deployment. Comfortable translating messy, real-world data and complex AI systems into validated, stakeholder-ready solutions that both technical and non-technical audiences can act on.
Artificial Intelligence Engineering
Designing intelligent applications powered by machine learning, Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), AI agents, and modern AI frameworks.
Data Science & Machine Learning
Developing predictive models, performing data analysis, feature engineering, model evaluation, and transforming data into actionable insights.
Data Analytics & Visualization
Creating interactive dashboards, analyzing business performance, and communicating insights through effective data visualization and business intelligence tools.
Technical Skills
A comprehensive toolkit for building intelligent AI applications, machine learning solutions, and data-driven systems.
Programming Languages
AI & Machine Learning
AI & Data Frameworks
Data Analytics & Visualization
Databases
Cloud & DevOps
Featured Projects
A selection of AI, machine learning, and data science projects.
End-to-End ML Pipeline for 5G Network Anomaly Detection
Built a real-time anomaly detection pipeline that cuts alert volume by 85% while matching a realistic daily review capacity, deployed end-to-end with a live Kafka-to-dashboard serving system.
- • 85% Fewer Alerts vs Naive Rule
- • 2.6x Detection Rate vs Random Baseline
- • 994-Record Verified Replay
Real-Time Fraud & Risk Scoring System
Built a real-time fraud detection and risk scoring system for e-wallet transactions, using LightGBM and a Kafka streaming pipeline to flag fraud with 0.92 PR-AUC on a severely imbalanced dataset.
- • 6.3M Transactions Processed
- • 0.92 PR-AUC
- • $807K Net Value
Uplift-Based Marketing Optimization
Built an uplift model on 64K+ customer email campaign records to identify which customers converts because of an email, cutting send volume by 50%.
- • 4x Random Targeting
- • 50% Email Volume Cut
- • 64K+ Customer Experiment
Automaded X-Ray Threat Detection System
Built a real-time X-ray object detection system using YOLOv12 at ~60+ images per second to identify prohibited items.
- • 82.7% mAP@0.5
- • <16ms Inference Speed
- • 82.8% Precision
Experience & Education
My professional and academic journey in AI and data science.
Data Analyst Intern
Telkomsel, Enterprise Digital Service Delivery
- Cut manual data-integration effort by ~95% (from 1–2 hours to under 5 minutes per cycle) by designing and implementing a unified consolidation workflow for ~4,800 sales pipeline records sourced from Area Account Manager (AAM) and Corporate Account Manager (CAM) systems.
- Strengthened data reliability for enterprise decision-making by auditing and correcting data quality issues such as mismatched IDs, incorrect date formats, missing fields, and SLA formula errors, then presenting validated, analysis-ready datasets to 6 senior stakeholders.
- Enabled real-time visibility into revenue and delivery performance by building 3 interactive Looker Studio dashboards tracking revenue pipeline, project completion, and project trend metrics for enterprise digital solution reporting.
Get In Touch
I'm always interested in new opportunities and collaborations. Let's discuss how we can work together.