Miftahul Anwar Email & Phone Number

Jakarta Selatan, Jakarta, Indonesia

- Revamped zenius.net’s navbar. During a month-long A/B test, user interaction with content increased by +38.44%. Utilizing Next.js and MoEngage.- Integrated Keycloak for ZenProX authentication. Also crafted custom UI templates using Keycloak's FTL- Implemented Next-i18n to enable app translation between English and Indonesia, using Next.js.- Streamlined OpenAI fine-tuned model management with a web-based panel and reporting system,simplifying tutor assets conversion. Using Next.js, NestJS, Tailwind CSS, Firebase, and OpenAI API.- Worked with the team to gather and refine tutor-provided subject problem assets, creating a fine-tuned AI model to assist in problem generation based on those assets. Utilized Python and OpenAI API.- Contributed to the development of a web-based AI Assistant for generating subject problems based on tutor assets, utilizing Next.js, NestJS, Tailwind CSS, Firebase, and the OpenAI API.

- Contributed to a Clojure-based Problem Generator engine, facilitating the creation of diverse schoolsubject problems, and leading to 11.9K monthly active users and over 1M questions answered weekly.This system streamlined content creation from tutor templates to Zenius App-ready JSON files.- Developed a Clojure-based app that assists tutors in quickly generating quizzes, infographics, and SVG graphs. Streamlining the creation of educational content.

South Jakarta

- Collaborated on a project to reverse engineer an embedded device, with the goal of deepening my understanding of its functionality and performance. Utilize Zephyr RTOS for sensory access and perform related tasks- Developed a neural network model to identify GPS Jammer signals, distinguishing between malicious jamming or common signal loss, using TensorFlow and Python- Created an Android app to run the GPS Jammer neural network model, sending real-time notifications to differentiate between malicious jamming or common signal, using Kotlin and MQTT

Surabaya

- Collaborated with the team to develop the IoT device that is capable to assess human health conditions (ear, lungs, heart, mouth, skin, and body temperature), incorporating efficient sensory access for accurate data collection and transmission to a web server, using C++ and Python- Developed a user-friendly Graphical User Interface for seamless user interaction and effective device communication, enabling accurate data analysis and display, using Qt framework on C++

- Led the management of research activities, working closely with team members to coordinate tasks and ensure smooth project progression while addressing any challenges- Carefully developed a comprehensive research roadmap to guide the team's efforts, maintain focus on key objectives, and facilitate efficient progress towards the desired outcomes- Prepared a conference paper, summarizing research findings, incorporating data analysis, and employing clear communication to effectively convey the project's outcomes and insights

- Developed a Graphical User Interface (GUI) to wirelessly monitor and analyze humanoid robot soccer player behavior, leading to improved gameplay, better analysis, and optimization of performance. Utilized the Qt framework and C++ for development- Collaborated with a team to refine the decision-making algorithm for humanoid robot soccer players, resulting in enhancements to ball-finding algorithms, ball approaches, path planning, and individual attacking strategies, increasing performance and efficiency during gameplay- Worked collaboratively to upgrade the robot teamwork algorithm, leading to improved role allocation and teamwork performance for humanoid robot soccer players- Designed a robot localization system, contributing to better teamwork and individual performance of humanoid robot soccer players through more accurate ball tracking, efficient attacking and defending strategies, and precise positioning. Implemented using the Particle Filter algorithm- Developed an image processing algorithm, enabling accurate identification of goalpost, landmarks, and the ball, and enhancing the efficiency and effectiveness of the robot localization system for humanoid robot soccer players. Utilized MobileNetV2 for the solution