AI solution architecture
Turning business requirements into end-to-end AI architectures — data, models, deployment and governance — alongside global, cross-functional teams.
Solution designEnterprise AIGovernance
Complex problems.Practical AI.
Hi, I’m Marvel Martawidjaja — an AI solution architect and data scientist. I connect business needs with intelligent systems, from enterprise AI architecture and applied research to machine learning in production.
AI Solution ArchitectPhilip Morris International
B.Sc. Data ScienceBINUS University · 2026
Experience across industry & research
01 / What I do
Where business needs meet intelligent systems.
Three things I bring to a team — whether that’s a global enterprise, a financial regulator, or a research lab.
Machine learning in production
NLP, recommendation and predictive systems built to run — Databricks pipelines, MLflow deployment, monitoring and drift detection on real customer data.
NLPMLOpsDatabricks
Applied research
Peer-reviewed work across healthcare, climate, energy and computer vision — with a focus on models that are explainable and trustworthy.
Explainable AIComputer visionDeep learning
02 / Experience
From research
to real-world impact.
Working across enterprise AI, financial services, heavy industry and academia to turn data into useful decisions.
My full experience
03 / Beyond the day job
Teaching, speaking & competing.
NVIDIA · Guest speaker
National press briefing
One of three students selected to present AI innovation alongside NVIDIA leaders to national tech media.
Stanford · Teaching assistant
Data science short course
Supported 120+ students with Prof. Jennifer Widom, Dean of Stanford Engineering.
NVIDIA DLI · Teaching assistant
LLM fine-tuning workshop
Guided 50+ participants through PEFT and LLM customisation with NVIDIA NeMo.
PEDAS · 2nd runner-up
National data competition
Placed third nationally at Pesta Data Nasional, hosted by UIN Syarif Hidayatullah.
Let’s connect