Master of Actuarial Science
Oct, 2016Bandung Institute of Technology
Quantitative risk management and predictive methods, including research on IBNR, ruin probabilities, and ultimate loss estimations.
Resume / Professional record
Aryanto, M.Si
MLOps Architect for Production ML/GenAI · Recommender Systems · Generative AI · Computer Vision · Agentic AI
A career spanning data analysis, data engineering, data science, quantitative modeling, machine learning, computer vision, AI applications and technical leadership.
9+
Python/R/Cython/C++ programming years
7+
ML/AI Systems Developed years
9+
structured/nonstructured DB years
20+
people mentored
Experience map
Career by formal role
Formal titles are kept intact. Discipline tags display adjacent technical areas for each role. Some engagements overlap as reflected in the primary career history.
Data Science Manager
Merkle Indonesia · Dentsu|Mar 2022 – Dec 2025 · 3 yrs 10 mos
Summary: Led end-to-end data science, generative AI, and advanced analytics delivery across enterprise automotive, telco, banking, insurance, and e-commerce clients.
- Architected Agentic AI workflows for automated information extraction from unstructured PDFs, Excel files, and contracts, boosting data processing efficiency by 40%.
- Engineered enterprise LLM conversational AI chatbots integrated with Gemini, Dialogflow, and RAG knowledge bases, achieving >85% resolution accuracy for financial/insurance clients.
- Deployed Next Best Offer (NBO) and cross-sell recommendation engines across retail and telco verticals, driving $5.84M+ in aggregate influenced revenue.
- Designed Learning-to-Rank (LTR) algorithms and artificial rating engines for e-commerce products lacking native user reviews, increasing cross-category conversions by 18%.
- Optimized Customer Data Platform (CDP) and Gold Layer data pipelines on GCP BigQuery, unifying real-time event streams from GA4, Mixpanel, AppsFlyer, and Qualtrics.
- Implemented Multi-Touch Attribution (MTA) and predictive behavioral scoring models, optimizing marketing spend and improving client campaign ROI by 22%.
- Standardized production-grade MLOps best practices, high-performance SQL query structures, and maintainable Python design patterns across all client accounts.
- Managed and mentored a cross-functional team of 15+ Data Scientists and Analysts, maintaining 100% on-time delivery while translating executive requirements into AI solutions.
Data ScienceML EngineeringAnalyticsGenerative AIAgentic AILLMData PlatformLeadership
Senior Data Scientist
Asia Pulp & Paper|Jan 2021 – Mar 2022 · 1 yr 3 mos
Summary: Applied advanced machine learning, mathematical optimization, Agentic computer vision, and geospatial analysis to forestry operations, supply chain, and marketing.
- Engineered end-to-end computer vision and remote sensing deep learning pipelines (YOLO, ResNet) for precision forestry, tree recognition, and plant-health monitoring.
- Deployed multi-spectral satellite and drone image analysis models for real-time flood detection and blank spot identification across thousands of plantation hectares.
- Integrated Agentic document processing scripts to auto-parse geospatial metadata and environmental survey logs into centralized spatial databases.
- Constructed predictive yield forecasting and preventative maintenance models, reducing operational planning cycles by 30%.
- Formulated mathematical optimization and statistical models for forestry ROI estimation and macro-level community resource allocation.
- Modernized legacy data science systems by replacing ad-hoc script workflows with containerized, reproducible production pipelines.
- Led cross-functional deployments connecting ML model outputs directly to relational databases, executive dashboards, and mobile apps.
- Mentored junior data scientists on spatial feature engineering, algorithm selection, and high-performance Python/Cython code execution.
Data ScienceGenerative AIAgentic AIForecastingComputer VisionOptimizationGeospatialML
Data Scientist
Sinarmas Asia Pulp & Paper|Nov 2019 – Jan 2021 · 1 yr 3 mos
Summary: Combined consumer analytics, rigorous experimentation, and deep learning models with forestry remote sensing data.
- Processed multi-terabyte datasets using R, Python, PySpark, and SQL on distributed compute clusters for operational intelligence.
- Built deep learning object detection models for aerial forestry mapping, automating tree counting and canopy density evaluations.
- Engineered automated feature engineering pipelines and hyperparameter tuning routines (Optuna) for gradient boosting models (XGBoost, LightGBM).
- Developed time-series sales forecasting models for marketing and supply chain divisions, reducing inventory forecast variance by 15%.
- Designed and executed A/B testing protocols and statistical experimentation to validate algorithmic impact prior to production rollout.
- Constructed custom ETL pipelines integrating unstructured GIS raster images with structured relational databases.
- Created consumer behavior metrics, KPI analysis frameworks, and behavioral clustering to identify operational and consumer personas.
- Implemented automated data validation checks and communicated quantitative research insights directly to non-technical business stakeholders.
AnalyticsMLForecastingDeep LearningRemote SensingData Engineering
Data Manager
GSMA|Jul 2019 – Nov 2019 · 5 mos
Summary: Managed field survey architecture, telemetry validation, and predictive modeling for an off-grid solar electricity and e-money (LinkAja) adoption study in Sumba.
- Designed and deployed mobile ODK Collect survey instruments and managed central ODK Aggregate infrastructure for off-grid rural communities.
- Constructed automated data cleaning pipelines to deduplicate incoming household survey telemetry and revenue metrics.
- Performed rigorous descriptive and inferential statistical modeling to evaluate customer adoption drivers for e-money (LinkAja) electricity payments.
- Developed predictive machine learning models to identify socio-economic determinants influencing microgrid adoption and willingness-to-pay.
- Engineered geospatial mapping overlays connecting household survey respondents with local microgrid distribution nodes.
- Translated complex field research questions into structured quantitative reports and publication-ready technical documents for GSMA stakeholders.
- Provided technical guidance and daily operational supervision to field enumerator teams to minimize survey bias and logging errors.
- Partnered with cross-border research teams to align project outcomes with global sustainable development goals (SDGs).
Data ManagementStatisticsSurvey AnalyticsMLAnalytics
Data Engineer
AEConsults|Aug 2017 – Jul 2019 · 2 yrs
Summary: Built cloud data architectures, GIS mapping systems, automated ETL pipelines, and executive dashboards across multi-regional infrastructure projects.
- Architected end-to-end cloud data pipelines on Microsoft Azure using Azure SQL, Azure Data Factory, and Azure Functions.
- Built automated ETL workflows integrating structured relational databases with unstructured spatial data across remote infrastructure sites.
- Established robust CI/CD deployment workflows via GitHub Actions for automated data transfer and pipeline integration.
- Engineered custom GIS mapping databases to store, index, and query spatial topologies for regional utility planning in Papua and NTT.
- Developed real-time interactive logistics dashboards to monitor field equipment deployment and supply chain operations.
- Implemented RESTful APIs and CRUD database functions on Azure Web Services to support mobile field applications.
- Executed database performance tuning, indexing, and query optimization for high-frequency database read/write workloads.
- Enforced strict data governance, automated backup routines, and role-based access control policies across cloud data stores.
Data EngineeringETLGISAzureCI/CDCloud Architecture
Data Analyst
Electric Vine Industries|Jun 2017 – Jul 2019 · 2 yrs 2 mos
Summary: Built analytical datasets, predictive microgrid load models, and interactive BI dashboards to guide strategic and field operations.
- Designed and optimized MySQL relational databases to support high-throughput operational tracking and customer telemetry.
- Developed interactive Power BI and Looker Studio dashboards delivering executive visibility into microgrid performance metrics.
- Formulated predictive models to forecast customer energy demand, optimizing battery storage and solar microgrid distribution.
- Conducted cohort analyses and behavioral customer segmentation using Python, R, and SQL to identify energy consumption trends.
- Standardized reporting workflows into automated scripts, reducing manual reporting workload by 80%.
- Analyzed battery lifecycle telemetry to guide proactive preventative maintenance and reduce field downtime.
- Authored publication-grade technical and financial feasibility reports using LaTeX for internal executives and external investors.
- Facilitated field FGDs and translated complex stakeholder requests into clean SQL queries and structured diagnostic feeds.
Data AnalyticsDashboardBISQLPythonRML
Operation Data Analyst
Boltech.io (formerly AmTrust Mobile Solutions)|Jul 2016 – Jun 2017 · 1 yr
Summary: Applied statistical sampling, causal inference, and time-series forecasting to optimize mobile warranty product operations and CRM workflows.
- Mined enterprise CRM and operational databases using R and Python to identify operational bottlenecks and cost-saving opportunities.
- Built causal modeling frameworks to identify key operational drivers behind claim processing delays, improving SLA compliance by 15%.
- Formulated time-series forecasting models to project future customer claim volumes and align staffing requirements.
- Developed automated daily, weekly, monthly, and annual KPI reporting suites for regional executive management.
- Applied targeted customer segmentation techniques to boost mobile warranty renewal rates and cross-sell campaigns.
- Conducted statistical sampling and hypothesis testing to validate quality assurance standards across customer service touchpoints.
- Optimized core SQL queries and database feeds powering internal operational dashboards, accelerating report load times.
- Established standardized data dictionaries and metric definitions across cross-functional operations and product teams.
AnalyticsForecastingCausal AnalysisReportingData Science
Quantitative Risk Analyst
PT. LAPI ITB|Aug 2015 – Jul 2016 · 1 yr
Summary: Formulated and validated mathematical prediction models for Chevron Pacific Indonesia soil-remediation environmental projects.
- Formulated advanced quantitative risk models to estimate soil remediation volumes for Chevron Pacific Indonesia environmental projects.
- Developed mathematical correction factor models to account for excavation expansion and clean soil carryover dynamics.
- Built predictive statistical models to delineate actual soil volumes where Total Petroleum Hydrocarbon (TPH) exceeded 1%.
- Executed Monte Carlo computer simulations to stress-test and validate model behavior under diverse geological conditions.
- Translated complex mathematical formulas into executable C/C++ and Python simulation code.
- Conducted sensitivity and uncertainty analyses to establish reliable confidence intervals for volume estimations.
- Co-authored technical validation reports for regulatory compliance and executive review by Chevron stakeholders.
- Presented mathematical models to senior engineering consultants and implemented automated model monitoring checks in the field.
Quantitative ModelingForecastingRiskStatisticsSimulation
Freelance Data Scientist
Upwork|Jun 2013 – Mar 2016 · 2 yrs 10 mos
Summary: Delivered end-to-end data science, machine learning algorithms, custom web scrapers, and executive dashboards for international business clients.
- Executed global data science engagements, maintaining a top-rated track record across international client accounts.
- Designed diagnostic and predictive analytics roadmaps for e-commerce, healthcare, and financial tech clients.
- Implemented supervised and unsupervised machine learning models (XGBoost, Random Forest, K-Means) to drive client business goals.
- Built interactive web dashboards (Streamlit, Dash, Power BI) to accelerate real-time executive decision-making.
- Constructed custom time-series forecasting models for sales forecasting, inventory planning, and financial trend analysis.
- Constructed automated data scrapers and feature engineering pipelines for unstructured web datasets.
- Optimized mathematical algorithms in Python and C to accelerate model training and execution speed.
- Delivered clean technical documentation, code repositories, and production-ready software solutions for client engineering teams.
Data ScienceMLForecastingAnalyticsDashboards
Skill duration
Detail Experiences in Years
App Productionized using Python+9 years
DB structure [SQL/noSQL/VectorDB]+9 years
AI / ML / DL / RL+7 years
R for statistical analysis+7 years
Cython / C / C++ algorithm optimization+6 years
Computer Vision+5 years
NLP, LLM & Generative AI+5 years
Julia for backend+4 years
Javascript/Node.js+3 years
Selected projects
Work that connects data to outcomes
01Customer Data Platform & Gold Layer Analytics
02Behavioral Scoring & Lead Generation Models
03Next Best Offer & Cross-Selling Engines
04Agentic AI Document Extraction (PDF/Docs/Excel)
05LLM Conversational Chatbots with RAG Knowledge Bases
06Deep Learning Session & Product Recommendation
07Learning-to-Rank Product Recommendation Systems
08Large-Scale GA4 & Firebase Event Analytics Pipelines
09Multi-Touch Attribution Modeling on GCP BigQuery
Education
Mathematics
Mar, 2012Cenderawasih University
Applied mathematics with research exposure in numerical programming, modeling, and statistical analysis.
Professional signals
$5.84M+
business value influenced
Customer targeting, recommendation systems, operational optimization, Agentic AI, and LLM automation.
15+
people mentored
Data Science and Data Analyst teams.
10+
industry contexts
Automotive, telco, banking, insurance, e-commerce, forestry, microgrid energy, and more.
End-to-end capability
From data architecture, Agentic AI document processing, LLMs, and ML modeling to production systems, GCP data platforms, and executive team leadership.
Communication
Indonesian as native language. English for professional working proficiency, technical documentation, and executive collaboration.