Hi,
I'm Natasha
an ML Engineer
London, UK | (+44) 7470632088

Email Hire Me What I Build

I specialize in architecting Data Science and Machine Learning solutions. My expertise lies in developing predictive systems for fraud detection, customer behavioral analytics, and geospatial intelligence, as well as computational modeling and implementing LLM frameworks.

Outside of work, I usually spend my time writing, gaming, tinkering with IoT/3D modelling, or do strength training.


# ABOUT ME

I am a Graduate Researcher at Imperial College London, finishing my MSc in Applied Computational Science & Engineering.

I am currently seeking Senior Data Science or Applied AI roles (September 2026 start), looking for Remote (Global/APAC) opportunities based in Indonesia.

My technical core focuses on predictive modeling, end-to-end ML pipelines, and cloud-native deployments. I specialize in bridging the gap between complex data analysis and production-ready AI solutions.

# SHOWCASE



HiringHelp Chatbot (RAG LangChain)

View Project


Fraud Analytics with Graph Neural Network (GNN) - Syndicate Merchant Detection

View Project


Face Recognition with Arduino (OpenCV)

View Project
Capybara Pet Game screenshot

Capybara Pet Game (for funsies)

View Project


NLP Youtube Recommendation System (GCP OAuth 2.0 API)

View Project


Predictive Analytics (Machine Learning for Sales Forecasting)

View Project


AI Pickup Line Generator (smolLM)

View Project
Spaced Repetition Learning App screenshot

Spaced Repetition Learning App (Python)

View Project


Marketplace Web Page (Vue.js & Go)

View Project

# WHAT I BUILD

ML systems, end-to-end.
From messy data to containerised APIs in production.

  fraud-detection   llm-rag   geospatial-intel   predictive-modelling   fastapi-docker   gcp-bigquery   graph-neural-nets   sph-simulation
↓ see my work experience

WORK EXPERIENCE

Research collaboration between Imperial College London & Research Fellow from NASA Ames Research Center.

● Conducted 203 SPH simulations (resolutions up to 10^7 particles) for planetary modeling Martian moon formation relevant to the MMX mission, identified ~47% mass retention and ~52% escape trajectories in alignment with Kegerreis (2024).

● Developed predictive machine learning frameworks to process billions of data points by mapping initial conditions to stabilized 90ks end-states snapshots, implementing Friends-of-Friends (FoF) clustering algorithms for statistical data analysis.

● Leveraged High Performance Computing (HPC) on the Imperial CX3 cluster to scale and parallelize SWIFT simulations, interpreting complex 3D formation dynamics for large-scale physical system modeling.

Dans is an IT solutions provider for Telkomsel (Indonesia's No.1 internet provider, 150M+ users), the company behind DigiPOS and AI-powered Digital Smart Care.

● Led development of a Fraud Detection System, applying and benchmarking 6+ anomaly detection methods to deliver real-time fraud scoring models.

● Developed practical ML solutions in 5 projects for Telkomsel, including digital point-of-sale and customer service chatbots.

● Designed and implemented scalable AI architectures combining rule-based logic and ML inference, achieving a 15% increase in precision.

● Implemented containerized FastAPI services for a real-time AI monitoring platform, incorporating ML-based alerting, geospatial, and RAG.

● Stack: Python, GCP, BigQuery, VertexAI, MLflow, Dataflow, Apache Kafka, FastAPI, Docker

LinkAja is Indonesia's largest state-owned digital wallet (93M+ users, 2.9M merchants, 100%+ YoY GTV growth).

● Led development of 4 ML fraud detection models and streamlined BI dashboard monitoring for B2B & B2C e-wallet fintech.

● Mentored new engineers on fraud analytics and AI/ML techniques via MSIB program.

● Prevented 95% of potential monthly financial losses & reduced fraud incidents by 20.3% compared to the previous model.

● Closely collaborated with 3 teams to deploy ML models on GCP Cloud Functions and VMs.

● Stack: Python, GCP, BigQuery, Cloud Functions, Virtual Machines

LAPAN is Indonesia's state research agency responsible for advancing national R&D in AI, satellite technology, and aerospace. (Now integrated into BRIN, National Research and Innovation Agency)

Research Associate, AI/ML
● Applied machine learning and data analytics to satellite-based climate and atmospheric data under a state research grant.
● Co-authored two publications in collaboration with Bandung Institute of Technology (ITB), Indonesia's top-ranked university.

Data Science Research Intern
● Forecasted monsoonal index (1965-2022) with ARIMA in Python & R at 95% accuracy; visualized using GeoPandas and Seaborn.
● Modeled data from 900+ weather stations, producing insights for LAPAN's further research.

Education

  • IMPERIAL COLLEGE LONDON
     MSc in Applied Computational Science and Engineering
     ~ Ranked #2 in the World (QS World University Rankings 2025 & 2026), awarded the fully funded LPDP Scholarship


  • UNIVERSITY OF INDONESIA
     Postgraduate Studies in Environmental Science
    (Energy & Sustainability)
     ~ Ranked #1 in Indonesia (Times Higher Education 2024-2026 & QS WUR 2023-2025), prior graduate coursework


  • BANDUNG INSTITUTE OF TECHNOLOGY
     BSc in Earth Science and Technology, Meteorology
    (Geospatial Data Science)
     ~ Top #1 for Engineering & Technology in Indonesia (QS WUR 2023 & 2025-2026)

TECH STACKS
Programming Languages:Python, C++, Javascript
Libraries/Tools:ML (Scikit-learn, TensorFlow, PyTorch, Keras, XGBoost, OpenCV)
NLP (NLTK, HuggingFace, spaCy)
Visualization (Matplotlib, Seaborn, Pandas, NumPy, Plotly)
Graph Networks (NetworkX, RGCNconv, HGTconv)
Generative AI (GPT-4, Llama, Stable Diffusion/SDXL)
LLM Framework (OpenAI, LangChain), Vertex AI, RAG (Retrieval-Augmented Generation)
Databases:PostgreSQL, BigQuery, MySQL, DBMS (DBeaver)
Development:Code Editor (VS Code, Google Colab), Version Control (Git), Frontend Frameworks (Vue.js, React.js), Backend Frameworks (Flask, FastAPI, Django)
Deployment:Compute Engine VM, Cloud Functions, Docker, Kubernetes


scroll back to top