MRP

Made Raditya Pujamurti

I help companies improve efficiency by deploying workflow automation, data processing and AI models

AI Engineer at hyrd GmbH

Germany

radityapujamurti[at]gmail[dot]com


Stats

Team Work

Problem Solving

Coding

Curiosity


Tech Stack

Python | LangChain | Pytorch | Javascript | MongoDB | n8n | Elasticsearch Logstash Kibana (ELK) | SQL | Tableau | Power BI | Unix


Languages

Indonesian

Native language

English

Fluent

German

A1


Work Experience

AI Engineer | hyrd GmbH
August 2024 - Present
  • Build LLM powered agents.

Machine Learning Engineer | Alstin Capital
April 2024 - July 2024
  • Build and deploy startup recommendation systems using custom machine learning models.
  • Boosted investment analysts' productivity by recommending selected promising startups.

Data Science Werkstudent | Alstin Capital
June 2022 - March 2024
  • Developed and implemented a robust ETL (Extract, Transform, Load) automation pipeline for processing vast volumes of startup data, eliminating time-consuming manual tasks and accelerating data ingestion and transformation processes.
  • Successfully built and deployed a smart filtering system using a machine learning model to optimize lead conversion. This model led to a remarkable improvement in conversion rates, exceeding 50% above previous benchmarks.

Database Administrator | Garena Ltd
July 2017 - August 2020
  • Proficiently deployed, managed, and maintained an extensive infrastructure of MongoDB servers, totaling in the hundreds. This involved ensuring the availability, performance, and security of these critical data storage systems.
  • Successfully implemented a robust backup system to safeguard against data loss. This meticulous backup strategy ensured that critical data was regularly and securely archived, reducing the risk of significant data loss due to unforeseen events.
  • Introduced a comprehensive monitoring and alert system for the MongoDB servers, enabling real-time identification of potential issues and immediate response to anomalies. This proactive approach enhanced system reliability and minimized downtimes.
  • Automated the deployment and management of MongoDB servers to streamline operations, reduce manual errors, and enhance efficiency. This automation significantly improved the agility of the database infrastructure, allowing for rapid scalability and reduced operational overhead.

Web Developer Intern | Optimai Pte Ltd
January 2016 - June 2016
  • Built front end of a .NET web application to screen and monitor stocks and derivative products.
  • Performed User Acceptance Testing (UAT) which led to successful product launch.
  • Optimized MSSQL DB operation.

Projects

E-commerce Data Analytics & Interactive Dashboard
Data Analytics EDA Data Cleaning Python Pandas Tableau

This project includes an interactive dashboard using Tableau. You can change various parameters such as moving average period and top N sales/expenses. When certain period is selected, other dashboard components will be adjusted automatically. Exploratory data analysis (EDA) and data cleaning were also performed to ensure more accurate results.


Conversational German Practice Web App
Machine Learning NLP Transformer Model Python Javascript FastAPI Svelte

Used a local large language model (LLM) to create a bot that can respond appropriately to a given verbal input. To keep it cost effective, no paid service such as OpenAI API is used. Built using Svelte for frontend and FastAPI for backend.


AI Research Agent
LLM Python FastAPI Nginx Linux

Created an AI agent that is able to perform research for a certain task using GPT and Langchain. Deployed the agent on a cloud server with FastAPI, Nginx.


Low-Code Workflow Automation
Data Engineering Javascript ETL n8n

Built automation workflows using a low-code tool called n8n to perform extract transform load (ETL) process. Integrated the workflows with external services such as CRM, API and database.


Lifelog Image Smart Search
Computer Vision Python Transformer Model

Leveraged LLMs to perform quick search of tens of thousands lifelogs of images. Combined image and text embeddings to produce accurate images recommendation for a given query.


Trading Strategy Tester
Data Analysis Python Streamlit Docker

Built back testing web application using Streamlit to allow users to test trading strategy without coding.


Education

Universität Koblenz
October 2020 - November 2023

Master Degree | Web and Data Science


National University of Singapore (NUS)
August 2013 - August 2017

Bachelor Degree | Computer Engineering


Leadership Experience

President | NUS Indonesian Student Community (PINUS)
July 2015 – August 2016

Head of Welfare Division | NUS Indonesian Student Community (PINUS)
May 2014 – July 2015

Project Executive Head | NUS Social Entrepreneurship Experience and Development (SEED)
January 2014 – August 2014

Vice-President | SAJC 34th Student Council
May 2011 – May 2012