Current position
Computer Science undergraduate at Universitas Indonesia since 2022. Currently working on my thesis at LEXIN.
Academic baseCS @ Universitas Indonesia
Mariano Gerardus Senduk
Guided by the North Star of AGI
Computer Science student interested in how intelligence works and how to turn that understanding into systems that help people. I work across AI research and product engineering, with AGI as my long-term direction.
About
Computer Science undergraduate at Universitas Indonesia since 2022. Currently working on my thesis at LEXIN.
Academic baseA foundation in mathematics, algorithms, and machine learning, developed through coursework, research projects, and software engineering.
Technical foundationAI research and product engineering, alongside an interest in business, finance, and how technology fits into people's work.
Areas I exploreContributing to AGI by building useful, reliable, and understandable AI systems.
Long-term journeyProjects
Click a star to read the domain signal.
Compared lexical, hybrid, and dense retrieval with multilingual rerankers over 3.19 million documents, placing 4th on the class Kaggle leaderboard and studying ranking-quality versus runtime trade-offs.
Reinforcement LearningModeled pandemic response as an MDP in a stochastic four-city SEIRD Gymnasium environment, then compared PPO, A2C, and a restricted DQN baseline under health, economic, and budget constraints.
Deep Learning / FinanceDeveloped a causal multivariate forecasting model to predict 15-minute future log returns for 100 target stocks on the Indonesia Stock Exchange, using high-dimensional market features, Temporal Convolutional Networks, baseline comparisons, and calibration for near-zero market movement.
Information RetrievalExtended the inference procedure of ListT5, a listwise reranker based on Fusion-in-Decoder T5, by studying grouping strategies, tournament-style reranking, caching behavior, and BEIR benchmark trade-offs without retraining the underlying model.
Information RetrievalBuilt a complete mini search engine using Python, NumPy, SciPy, and FAISS, covering indexing, compression, TF-IDF, BM25 variants, WAND optimization, LSI retrieval, approximate nearest-neighbor search, and query evaluation.
Speech AIStudied detection and localization of synthetic segments in speech using Wav2Vec2-XLSR features and temporal attention, reporting both utterance-level and frame-level error rates.
Journey
Built the competitive foundation through informatics contests, algorithmic problem solving, and early computer science discipline.
Studying computer science while building projects across AI research, software systems, data, product work, and team-based engineering.
Worked with end-to-end ownership in a startup, understanding legal-product workflows and delivering features across AI, backend, data, and frontend systems.
Researching how LLMs can translate Indonesian judicial reasoning into readable symbolic representations while preserving the meaning of the source, at the Center for Legal Informatics, UI.
Connecting AI research, product work, digitalization projects, and business learning into a clearer long-term AGI trajectory.
Skills
North-east technical orbit
Builder orbit
Human workflow orbit
Market reasoning orbit
Self Blog