Athilla Zaidan, AI @GDP Labs

Software and AI engineer from Institut Teknologi Bandung. I build intelligent systems from model design to production interfaces.

EngineeringIntelligence

Profile

About

Proof-first systems, built to ship.

I work where machine learning, web systems, and operational tooling meet. The throughline is practical evidence: measurable model performance, resilient admin flows, and interfaces that make technical decisions legible.

01

Make intent observable.

I turn ambiguous product or model goals into screens, scores, logs, and workflows people can inspect.

02

Give information a structure.

Good systems reduce noise: domain rules, data movement, and user decisions should have a clear shape.

03

Keep craft and evidence together.

The interface has to feel considered, but the proof still comes from metrics, operations, and shipped behavior.

04

Build for real interpretation.

AI output, admin state, and risk signals only matter when they help someone understand what to do next.

Engineering range

Domain

01

Research and Applied AI

Anomaly detection, NLP classification, environmental risk scoring, and verification workflows shipped as usable systems.

Cognify, Vokara, WAVE, Virens
02

Fullstack Systems

Registration, payments, dashboards, file workflows, multi-tenant operations, and admin tools for real events.

FindMi, Grandsummit, Scresh
03

Low-Level Craft

Operating systems, graph algorithms, desktop apps, and backend foundations built close to the machine.

Malam KeOS ini, Mjolnir, Graph Theory
  • Python
  • Go
  • TypeScript
  • JavaScript
  • C
  • C++
  • Java
  • SQL
  • Haskell
  • HTML
  • CSS
  • PyTorch
  • Hugging Face
  • scikit-learn
  • NumPy
  • Pandas
  • SHAP
  • FastAPI
  • Jupyter
  • Next.js
  • React
  • React Native
  • Tailwind CSS
  • Drizzle ORM
  • Supabase
  • Docker
  • Git
  • Framer Motion
  • Python
  • Go
  • TypeScript
  • JavaScript
  • C
  • C++
  • Java
  • SQL
  • Haskell
  • HTML
  • CSS
  • PyTorch
  • Hugging Face
  • scikit-learn
  • NumPy
  • Pandas
  • SHAP
  • FastAPI
  • Jupyter
  • Next.js
  • React
  • React Native
  • Tailwind CSS
  • Drizzle ORM
  • Supabase
  • Docker
  • Git
  • Framer Motion

Selected projects

Works

Cognify project screenshot
AI Engineer

Cognify

Passive ADHD monitoring system using unsupervised learning on smartphone behavioral metadata. 3-layer anomaly detection with Z-Score, Isolation Forest, and SHAP-explainable scoring.

  • Python
  • scikit-learn
  • SHAP
  • NumPy
0.846 AUC-ROC · 100% Precision · 77.4% F1
WAVE project screenshot
AI & Fullstack Engineer

WAVE

AI-assisted water pollution monitoring platform with citizen photo reports, Sentinel-2 anomaly detection, OpenStreetMap source attribution, and explainable zone-level risk scoring.

  • Next.js
  • React
  • TypeScript
  • Python
Vokara project screenshot
AI & Backend Engineer

Vokara

NLP-based aspiration intelligence pipeline for policy prioritization. Fine-tuned IndoBERTweet for multi-task classification, clustering, and local LLM-based policy brief generation.

  • Python
  • FastAPI
  • PostgreSQL
  • HuggingFace
85.19% Macro F1 · 6 layer AI Pipeline
Scresh project screenshot
Fullstack & AI Engineer

Scresh

Multi-tenant cooperative digitalization platform for member financing, governance audits, and AI-powered vegetable freshness scanning with grade, confidence, and shelf-life estimates.

  • Next.js
  • React
  • TypeScript
  • PostgreSQL
TechnoScape 2026 Final
Virens project screenshot
Fullstack & ML Engineer

Virens

Nature-based carbon credit verification platform with project dossiers, satellite NDVI narration, anomaly tracking, verification scoring, and printable reports.

  • Next.js
  • TypeScript
  • Tailwind
  • Drizzle
Grandsummit 15th project screenshot
Head of IT, Fullstack

Grandsummit 15th

Registration platform for 200+ users with dual-track BCC/MCC flows, Supabase Auth/Postgres, dual-storage upload pipeline, and Google Sheets API sync.

  • Next.js
  • React
  • TypeScript
  • Tailwind
IT Aku Masuk ITB project screenshot
Fullstack Developer

IT Aku Masuk ITB

FindMi e-commerce platform (IDR 200M+ revenue) and AMI event platform (>10,000 users). Features Midtrans payments, QR ticketing, real-time scanners, and admin dashboards.

  • React
  • Vite
  • Hono
  • Drizzle
Mjolnir project screenshot
Fullstack Developer

Mjolnir

Digital library system with image search (PCA) and text recommendation (LSA). Go backend with React + Vite frontend, Docker deployment.

  • Go
  • React
  • TypeScript
  • Tailwind
Graph Theory Visualizer project screenshot
Algorithm & Systems Engineer

Graph Theory Visualizer

Interactive graph theory visualizer with Next.js UI, 2D/3D graph views, C++ graph engine, TypeScript fallbacks, and demos for traversal, shortest path, MST, TSP, matching, and bandwidth algorithms.

  • C++
  • TypeScript
  • Next.js
  • Three.js
Malam KeOS ini project screenshot
Systems Developer

Malam KeOS ini

Custom 32-bit x86 operating system from scratch. Features Unix-like shell, EXT2 filesystem, round-robin multitasking scheduler, VGA graphics, and Bad Apple ASCII player.

  • C
  • Assembly
  • Makefile
  • QEMU
Pr
Desktop Developer

Proxivity

JavaFX fitness tracking desktop app with food logging, activity tracking, calorie targets, 7-day trend charts, PDF report export, and AI Healthbot Assistant.

  • Java
  • JavaFX
  • Maven
  • SQLite
Gr
Algorithm Engineer

Greedy Conqueror

Greedy-based bot implementation for Battlecode 2025. Three bot variants with scoring-based nearest-target movement, tower-first expansion, and coverage-maximizing strategies.

  • Java
  • Gradle
  • Battlecode
8th Place at Battlecode

Operating history

Experience

Jun. 2026 – Present

Artificial Intelligence Engineer Intern

GDP Labs

  • Delivered PRs across GLAIR's AI platform and Model Accuracy Benchmark stack, adding Claude Sonnet 5 adaptive reasoning and hardening OpenAI/Anthropic provider compatibility, response parsing, authentication, and local inference execution
  • Implemented streaming time-to-first-token measurement for title-generation benchmarks by integrating an SDK EventEmitter into the LLM invoker while preserving end-to-end latency tracking
Oct. 2025 – Mar. 2026

Technology Development Intern

Inkubator IT

  • Owned client communication and delivered a polished Next.js, React, TypeScript, and Tailwind CSS platform from complex Figma designs with glassmorphism UI and Framer Motion animations
  • Built a Brand DNA Quiz system with React Context, weighted scoring, custom SVG radar charting, and instant multi-axis personality profiling
Feb. 2026 – Present

Fullstack Developer and IT Lead

StudentxCEOs Grandsummit 15th

  • Led a 7-person dev and UI/UX team to ship sxcgrandsummit.com for Indonesia's premier student business competition
  • Architected a dual-track BCC/MCC registration system handling 280+ teams, 500+ participants, 1,500+ file uploads, Rp20M+ revenue, and real-time preliminary submission monitoring
Oct. 2025 – May 2026

Frontend Developer

TEDxITB 9.0

  • Built the TEDxITB 9.0 frontend using Next.js, React, TypeScript, and Tailwind CSS, leveraging App Router Server Components for SEO and Client Components for interactive commerce flows
  • Crafted interactive event experiences with Framer Motion, animated carousels, countdowns, responsive hero sections, admin order tables, OpenGraph metadata, and programmatic sitemaps
Oct. 2025 – Feb. 2026

Fullstack Developer

IT Aku Masuk ITB 2026

  • Engineered FindMi, a fullstack merchandise e-commerce platform using React, Vite, Hono, Drizzle, and PostgreSQL that generated IDR 200M+ revenue
  • Architected the Aku Masuk ITB event platform serving 10,000+ users across RSVP, multi-participant registration, barcode generation, and offline-resilient ticket display

Build something real

Contact

For AI engineering and fullstack systems.

Reach me for technical collaboration, internships, or project conversations. Source, profile, and resume are one step away.

GitHubLinkedInDownload CV