Available for Hire
Hi, I'm
Rafi Atha
Machine Learning Engineer &
Full-Stack Developer
Results-driven Machine Learning Engineer and Full-Stack Developer with experience building production-ready AI systems. Skilled in end-to-end delivery across backend, frontend, and DevOps. Worked on integrating AI into ERP workflows, including supporting the transition from Odoo to SAP, to improve automation and decision-making. Passionate about combining enterprise systems and AI to optimize business processes and help organizations work smarter.
Professional Experience
A timeline of my professional journey in Machine Learning, ERP Systems, and Full-Stack Development.
Machine Learning Engineer
TridorianFull Time
Apr 2025 — Present
- Developed a full multi-agent AI system using Vertex AI to automate operations and support decision-making at Tridorian.
- Built chat-based intelligent agents using ADK for context-aware interactions and dynamic workflow automation.
- Maintained and optimized multi-agent orchestration systems, ensuring scalability, performance, and alignment with business objectives.
AI Engineer
WalletYaPart Time
Apr 2025 — Present
- Built production-ready AI services for receipt-to-text and text-to-receipt OCR, enabling accurate and automated financial data extraction.
- Researched and fine-tuned OCR and LLM models to enhance recognition accuracy and contextual understanding of Indonesian receipts.
- Designed and maintained AI pipelines and APIs integrated with WalletYa’s backend, improving automation, reliability, and user experience.
ERP Software Engineer (L2)
Paragon CorpContract
Dec 2024 — Mar 2025
- Developed and enhanced ERP features to streamline manufacturing, distribution, sales, finance, accounting, and tax processes, ensuring alignment with current business operations.
- Integrated multiple ERP systems with surrounding applications using n8n and ni-fi.
- Implemented an AI-powered code review tool to reduce code review time by ~50%.
- Supported ERP transformation from Odoo to SAP including data migration, process optimization, and user adoption.
●SYSTEM_LOG: ACTIVE_PROJECTS
Neural Archives
// IDX_01
Embed AI in Google Workspace
2025
Machine Learning EngineerAI AgentAutomation
- ›Embedded the Dorian multi-agent orchestrator into Google Workspace using Agent Development Kit (ADK).
- ›Used an agent-to-agent (A2A) protocol to coordinate downstream agents for end-to-end deal execution.
Vertex AIPythonGoogle Workspace APICloud Functions+1
// IDX_02
Dorian
2025
ML EngineerChatbotNLP
- ›Built Dorian as the central orchestrator for Tria, Chrono, Atlas, and the AI Scoring Agent.
- ›Implemented agent routing and coordination using ADK with an A2A protocol for multi-step workflows.
ADKPythonGolang
// IDX_03
AI Scoring Agent
2025
AI EngineerOCRFinTech
- ›Developed an AI agent to prioritize customer deals using multi-parameter scoring.
- ›Analyzed meeting transcripts, deal value, scope, timeline, and supporting artifacts (Drive, email, and documents) to generate a priority score.
PyTorchFastAPIADK
// IDX_04
Chrono
2025
ML EngineerAgentOptimization
- ›Built an AI planning and forecasting agent to optimize engineer allocation and delivery timelines.
- ›Combined Jira + Tempo utilization data with deal artifacts (meeting notes, scope, Drive docs, and emails) to produce resource availability forecasts.
PythonVertex AIADK
// IDX_05
Tria
2025
ML EngineerClassificationRouting
- ›Built a marketing-facing website assistant for Tridorian to handle booking inquiries and guide users across the site.
- ›Helped users find information and complete newsletter/event subscriptions through conversational flows.
PythonADK
// IDX_06
Atlas
2025
ML EngineerRAGKnowledge Base
- ›Developed an AI sales enablement agent to extract SPICE deal signals from meetings, emails, and Google Drive documents.
- ›Translated deal understanding into structured SOW drafts with clear scope, assumptions, risks, and deliverables.
GolangADKPython
Engineered Solutions
A collection of deployed applications, ERP systems, and internal tools built for scale and operational efficiency.
Tricoin
DeveloperWeb3Personal Project
- A cryptocurrency portfolio tracker and market analysis dashboard.
- Integrated with public crypto APIs to fetch real-time pricing and market cap data.
Built With
ReactNode.jsCoinGecko API
Human Resource Application
Frontend DeveloperInternal ToolFrontend
- Contributed to the development of an internal HR tool for task management and employee performance tracking.
- Optimized workflows to streamline daily HR operations.
Built With
Vue.jsJavaScriptCSS
Inventory Management System
ERP OperationERPBackend
- Maintained and enhanced Odoo-based inventory modules for a large-scale manufacturing stream.
- Optimized database queries to handle high-volume stock movement data.
Built With
OdooPythonPostgreSQL
Similarity Application Checker
Full Stack DeveloperToolingAnalysis
- Built a tool to detect code plagiarism and similarity across student assignments or internal codebases.
- Utilized AST (Abstract Syntax Tree) parsing for structural comparison rather than just text matching.
Built With
PythonReactFastAPI
Verified Credentials
Professional certifications and advanced coursework validating expertise in AI, Cloud Computing, and Machine Learning.
Google
2025Professional Machine Learning Engineer Certification
Google
2025Build with Vertex Technical Expert Badge
Google
2025Deploy Kubernetes Applications on Google Cloud
Google
2025Deploy a RAG application with vector search in Firestore
Google
2025Deploy an Agent with Agent Development Kit (ADK)
Google
2025Build search and recommendations applications with AI Applications
Google
2025Build Gen AI solutions using Model Garden models and APIs
Google
2025Analyze and Reason on Multimodal Data with Gemini
Stanford University (course)
2023Supervised Machine Learning: Regression and Classification
Stanford University (course)
2023Advance Learning Algorithms
Stanford University (course)
2023Unsupervised Learning, Recommenders, Reinforcement Learning
University of Alberta (course)
2023Fundamentals of Reinforcement Learning
University of Alberta (course)
2023Sample-based Learning Methods
Technical Arsenal
A robust stack of tools and frameworks, specializing in AI/ML orchestration and high-performance backend systems.
Python
Vertex AI
Gemini
OpenAI
LangChain
Agent Development Kit (ADK)
RAG
Prompt Engineering
PyTorch
TensorFlow
Scikit-learn
Pandas
NumPy
Hugging Face
Spoken Languages
Bahasa Indonesia (Native)English (Professional)
Academic Background
2020 — 2024
Universitas Indonesia
Bachelor Degree in Computer Science | Information Systems
- ▹Involved in organizations and committees (Sistem Fasilkom UI, Pesta Rakyat Komputer, COMPFEST, etc.).
Initialize Connection
Ready to collaborate? Send a signal for hiring, research inquiries, or project discussions.
user@contact-terminal: ~/send-message