Hi, I'm Jihad Akbar

Data Scientist

Specializing in Machine Learning, Python, SQL, Visualization, Deployment & Reporting to deliver data-driven solutions.

Get In Touch View Projects
Jihad Akbar

About Me

Personal Info

Career Summary

Data Scientist with hands-on experience in data analytics, engineering, and predictive modeling. Skilled in ETL pipelines, data preprocessing, and interactive dashboards. Certified in data science, analytics, and engineering, with expertise in Python, SQL, and visualization tools. Interested in applying machine learning for accurate predictions and delivering data-driven solutions.

Skills

Python SQL Machine Learning Deep Learning Time Series Forecasting Exploratory Data Analysis Data Visualization Tableau Power BI Streamlit Flask Ngrok GIT Docker TensorFlow PyTorch Hadoop HDFS LangChain FAISS RAG Airflow Kafka MLflow AWS Azure IBM

Languages

Indonesian
Native
English
IELTS 7.0 (C1)

Work Experience

Data Scientist

Oct 2025 - Present
PT Berca Hardayaperkasa
  • Built end-to-end ML pipelines including synthetic data generation, preprocessing, model training, and deployment with Medallion Architecture. Integrated Kafka for real-time streaming, Snowflake as data platform, and Streamlit dashboards for live, batch, and single predictions.
  • Developed production-grade Local RAG system with full n8n automation. Implemented intelligent routing for structured data (PostgreSQL) and unstructured data (Qdrant vector DB), semantic caching, conversation memory, Ollama for local LLM inference, and observability stack (Grafana/Prometheus/Loki) in Docker containerized environment.
  • Delivered technical pitch and presentation to client.

Machine Learning Annotator

Oct 2024 - Oct 2025
PT Nomura Research Institute Indonesia
  • Assessed AI model predictions by categorizing daily images and videos across multiple risk levels and evaluated outcomes to ensure model reliability and safety compliance.
  • Provided detailed feedback to the machine learning team, contributing to a significant improvement in model accuracy based on annotation insights.
  • Conducted precise image labeling with bounding boxes and segmentation, consistently achieving an annotation accuracy of 95%-100%.

Data Scientist

May 2025 - Aug 2025
PT Grand Delight International
  • Built an ETL pipeline to process internal data, extracting from Excel, preprocessing in Google Colab, and loading results via API into Google Sheets. Linked data to a Google Looker Studio dashboard for real-time sales and revenue insights.
  • Analyzed and visualized daily data to assess sales and service performance by villa, category, item, guest, etc. Delivered recommendations to improve revenue and optimize services, collaborating with the sales team to evaluate a month of realtime data for continuous improvements.

Project-Based Data Science

Feb 2025 - Mar 2025
Bukit Vista Hospitality Services
  • Developed an eye disease classifier for cataract, diabetic retinopathy, glaucoma, and normal conditions. Performed extensive image preprocessing, including checks for class distribution, image size, duplicates, corruption, blurriness, exposure issues, distortions, and noise, ensuring data quality for model training.
  • Built and fine-tuned deep learning models, selecting InceptionV3 with 83% accuracy as the optimal solution. Implemented techniques like early stopping and layer unfreezing for performance improvement, then deployed the model to production using Streamlit on the Hugging Face platform.

Annotation QA Lead

Feb 2024 - Sep 2024
CVAT.ai Corporation, Palo Alto, California, United States
  • Annotated diverse image datasets using CVAT’s tools to support advanced machine learning model development, ensuring high annotation accuracy and data quality.
  • Reclassified thousands of image objects into refined sub-classes for improved model training, applying bounding boxes and other tools for precise categorization.
  • Validated annotations from approximately 30 team members, achieving 90–100% accuracy by implementing quality assurance processes and providing feedback that boosted overall consistency and annotation standards across the team.

Data Project

Local RAG System with n8n

Built production-grade Local RAG System with n8n: dual-mode data handling (SQL + vector search), AI routing, semantic caching. Full-stack: Docker, web UI, 5 workflows, PostgreSQL, Qdrant, Elasticsearch, MinIO. Observability via Grafana/Prometheus/Loki, Ollama LLM inference.

Javascript n8n Ollama API RAG
View Project

Real-Time Network Incident Monitoring

Won 1st place Snowflake Hackathon 2025. Built Real-Time Network Incident Monitoring: Kafka→Snowpipe Streaming→Medallion Architecture. Predictive analytics for outages via Streamlit/Tableau dashboards. Integrated Cortex Analyst/Search, Intelligence chatbot, email automation in unified agent workflow.

Python Batching and Streaming Data Snowflake Kafka Streamlit Medallion Architecture
View Project

Full-Stack AI Chatbot for Credit Risk Assessment

Developed two AI chatbots for loan officers (Telegram) and borrowers (WhatsApp) to explore loan data, predict default risk, explain results, and export PDFs. Built dashboards for insights, automated workflows, and real-time data updates.

Python SHAP LangChain FAISS RAG OpenAI API HDFS PySpark
View Project

SmartDelivery: Zomato Operations Optimization

Predicted delivery times and improved driver efficiency with route analysis, visual dashboards, and recommendations for faster deliveries.

Python XGBoost OSRM API Nominatim API OpenWeatherMap API Streamlit Folium
View Project

FutureSight: E-Commerce Forecasting

Forecasted daily orders and revenue to improve inventory and revenue planning, with automated model updates and live dashboards.

Python ARIMA SARIMA Prophet Streamlit Google Sheets API
View Project

Eye Disease Classification

Built a deep learning model to detect cataract, diabetic retinopathy, glaucoma, and normal eyes, deployed for real-time diagnosis.

Python InceptionV3 TensorFlow Keras Streamlit Hugging Face
View Project

Churn Analysis

Identified customers likely to leave, uncovered key drivers, and provided business recommendations via an interactive web app.

Python XGBoost Streamlit
View Project

RFM Analysis for Customer Segmentation

Segmented customers in automobile sales to identify high-value clients and visualize trends through interactive dashboards.

Python Power BI
View Project

NYC Taxi Trip Duration

Predicted taxi trip durations and identified travel patterns using geospatial analysis and clustering.

Python scikit-learn LightGBM Folium
View Project

Credit Card Fraud Detection

Built a fraud detection model to flag suspicious transactions while reducing false alarms.

Python scikit-learn Extra Trees Classifier
View Project

Education, Certification, Achievement

Education

Data Science Bootcamp

Oct 2024 - Present
Dibimbing.id
  • Comprehensive program focused on data science and machine learning, covering essential skills in data analysis, data visualization (Tableau and Power BI), predictive modeling, and programming (Python and SQL).
  • Hands-on experience in real-world projects to build a robust data science portfolio.

BSc in Physics

Aug 2016 - Jul 2020
Physics Department of Universitas Sebelas Maret, Surakarta, Indonesia
  • Graduated Cum Laude, GPA: 3.75/4.00
  • Relevant Coursework: Statistical Physics, Physics Computation, Numerical Analysis, Physics-Based Modeling and Simulation

Certification

Data Science Bootcamp

dibimbing.id • May 2025

Python for Everybody Specialization

University of Michigan • Jan 2024

Deep Learning Specialization

DeepLearning.AI • Apr 2023

Machine Learning Specialization

Stanford University and DeepLearning.AI • Mar 2023

Achievement

1st Place Winner of Snowflake Hackathon Partner 2025

Nov 2025
  • Achieved 1st Place Winner in Snowflake’s Hackathon Partner 2025 by developing "Real-Time Network Incident Monitoring," an end-to-end platform that processes continuous network data streams from Kafka into Snowflake via Snowpipe Streaming. The system predicts potential outages in advance and supports both single and batch predictions, with live inference powered directly by the streaming data flow. Delivered real-time monitoring and trend analysis through Streamlit/Tableau dashboards.
  • Incorporated AI capabilities such as Cortex Analyst, Cortex Search, and Snowflake Intelligence (chatbot), along with a custom tool that sends chatbot-generated insight and recommendation summaries to email, all orchestrated within a single automated agent workflow. Built the solution on a Medallion Architecture (Bronze–Silver–Gold) to ensure clean, governed, and analytics-ready data across all layers.

Best Grade, MVP, Best Final Project, and Best Student of the Month at Data Science Bootcamp (dibimbing.id)

May 2025
  • Achieved The Best Grade (99.47, Grade A+), MVP, and The Best Final Project for "FutureSight: Revolutionizing E-Commerce with Data-Driven Forecasting" and "RFM Analysis for Customer Segmentation in Automobile Sales." Recognized as The Best Student of the Month by vote.

Get In Touch

Contact Information

Location

Jakarta, Indonesia

Email

jihadakbr@gmail.com

Send Me a Message