Applied AI Engineer · Researcher / Yogyakarta

Engineering intelligent systems.Building for production.

Muhammad Nabil Hanif. Applied AI engineer and researcher at the Computer Networks & Systems Laboratory (UII). Focused on reliable retrieval pipelines, applied machine learning, and high-performance web applications.

05Documented Awards
09+Shipped Projects
CN&SResearch Laboratory
2027B.Sc in Informatics

Software should be dependable, fast, and grounded in real data. I take ideas from concept to production.

  • 01 Clarity over complexity
  • 02 Verified in production
  • 03 Built for longevity
01

RAG and Knowledge Retrieval

Hybrid retrieval architectures combining dense semantic embeddings and BM25 sparse search for academic repositories and complex organizational documents.

  • + Hybrid BM25 and Dense
  • + Pinecone Vector Indexing
  • + Chunking Strategies
  • + Retrieval Evaluation
02

Real-Time Computer Vision

Touchless interactive systems and spatial interfaces driven by real-time hand landmark estimation and computer vision pipelines.

  • + MediaPipe Landmarker
  • + YOLOv8 Detection
  • + WebSocket Streams
  • + Low Latency Inference
03

Web Applications and Dashboards

Full-stack web applications with relational PostgreSQL schemas, role-based access control, geospatial analytics, and payment processing.

  • + Next.js
  • + TypeScript
  • + Supabase and PostgreSQL
  • + REST APIs
04

LLM Workflows and Automation

Production workflows connecting large language models to live databases, streaming inference endpoints, and structured data validation.

  • + Workflow Automation
  • + Fast Inference Endpoints
  • + Speech AI
  • + Structured Outputs
05

Mobile Development and Edge ML

Native mobile applications featuring on-device machine learning and optical character recognition for privacy and responsiveness.

  • + Kotlin and Android SDK
  • + Firebase Cloud
  • + ML Kit OCR
  • + On-Device Inference

04 / Selected Work

Selected projects.

01

PNDW Jogja MIS

Solo Lead Developer

Regional Public Complaint and Sentiment Intelligence System

A full-scale enterprise management information system for public complaints across the Yogyakarta Special Region. Features AI-powered sentiment analysis via Google Gemini, real-time executive dashboard with spatial mapping, role-based access control for 7 administrative tiers, and automated predictive report generation. Engineered solo from PostgreSQL schema design to production deployment.

Client Enterprise
Next.js 15TypeScriptSupabasePostgreSQLGoogle GeminiLeafletshadcn/ui
Recorded demo
02

IMUII Academic Intelligence

Undergraduate Final Project

Hybrid RAG Architecture for University Operations

A RAG-based cognitive assistant for university operations that answers complex academic queries grounded in verified institutional documentation. Hybrid retrieval pipeline combining dense semantic embeddings and BM25 sparse search over Indonesian-language corpora, Pinecone vector indexing, and GPT-based conversational memory backed by Supabase Realtime.

• Active Lab Research
Next.js 14TypeScriptN8NPineconeSupabaseOpenAI APIPython
Architecture SpecID // 02

IMUII Academic Intelligence focuses on decoupling indexing topologies from generation pipelines for high-throughput execution.

SYSTEM SPECYOGYAKARTA // ID
03

Vision Intelligence Constellation

Real-time Spatial Gesture Control & Spatial Computing

An interactive 3D tech constellation manipulated purely via real-time computer vision and hand landmarks. Depth-aware pinch to grasp nodes, two-finger rotation to orbit scenes, and zero-touch interface navigation. Combines Google MediaPipe landmarking with YOLOv8 object detection, optimized for Apple Silicon with adaptive frame-skipping at 30+ FPS over WebSockets.

Client Enterprise
PythonFastAPIMediaPipeYOLOv8OpenCVWebSocket
Architecture SpecID // 03

Vision Intelligence Constellation focuses on decoupling indexing topologies from generation pipelines for high-throughput execution.

SYSTEM SPECYOGYAKARTA // ID
04

Cuan AI

Tax & Financial Intelligence for 64M Indonesian SMEs

Specialized AI advisor built to navigate Indonesian SME taxation and cashflow strategy. Features an automated PPh Final tax calculation engine compliant with PP 55/2022 (including the Rp 500M tax-free threshold). Powered by Llama 3.3 70B via Groq with ultra-low latency streaming and deeply curated regulatory system prompts.

Next.js 16Groq APILlama 3.3 70BTypeScriptTailwind CSS
Recorded demo
05

DriveEase

Full-Stack Fleet Management & Car Rental Engine

A high-conversion car rental and fleet management platform tailored for tourism in Yogyakarta. Complete vehicle catalog with real-time specs, category filters, multi-step booking funnel, PostgreSQL transactions, and automated invoice reconciliation via Midtrans payment gateway.

Next.jsTypeScriptSupabasePostgreSQLMidtransTailwind CSS
Recorded demo
06

EcoSaver Mobile

Google Solution Challenge '25

AI & OCR-Powered Carbon & Consumption Tracker

Mobile sustainability application tracking household utilities and consumption habits. Features OCR receipt scanning, Gemini-powered analysis of purchase footprints, and customized carbon reduction recommendations. Built as technical lead for GDSC UII's entry in Google Solution Challenge 2025.

Client Enterprise
KotlinAndroid SDKFirebaseGemini APIML Kit OCR
Recorded demo
07

SkyCast Global

Dynamic Meteorological Forecasting Interface

Responsive weather application featuring dynamic diurnal color schemes that adjust in real-time to atmospheric metrics and solar azimuth. Provides drag-to-scroll hourly projections, 7-day outlooks, air quality telemetry, and wind direction radars.

Next.jsTypeScriptOpenWeather APIGeolocation APIFramer Motion
Recorded demo
08

secondYou

Automated Code Review & Static Security Audit

Developer productivity assistant providing granular AST analysis, vulnerability detection, and performance profiling for code diffs before production merges. Structured reports with severity indexing and remediation suggestions.

• Active Lab Research
Next.jsTypeScriptOpenAI APISupabaseTailwind CSS
Architecture SpecID // 08

secondYou focuses on decoupling indexing topologies from generation pipelines for high-throughput execution.

SYSTEM SPECYOGYAKARTA // ID
09

Blends

Speech-to-Insight Executive Meeting Intelligence

Meeting intelligence suite that transcribes multi-speaker conversations using OpenAI Whisper, extracts itemized action matrices per stakeholder, and runs chronological sentiment tracking across meeting recordings.

• Active Lab Research
Next.jsTypeScriptOpenAI WhisperOpenAI APISupabase
Architecture SpecID // 09

Blends focuses on decoupling indexing topologies from generation pipelines for high-throughput execution.

SYSTEM SPECYOGYAKARTA // ID

05 / About the Builder

Building reliable software with applied machine learning.

I am an independent software developer and applied AI engineer based in Yogyakarta, Indonesia. Currently completing my undergraduate studies in Informatics at Universitas Islam Indonesia (graduating 2027), I conduct research at the Computer Networks & Systems Laboratory (CN&S Lab UII).

My work centers on retrieval architectures and machine learning systems: grounding large language models in verified data, lowering latency, and reducing hallucination. From engineering solo public reporting platforms used across the Special Region of Yogyakarta to prototyping real-time computer vision interfaces, I focus on building dependable software for real-world environments.

AffiliationCN&S Research LabUniversitas Islam Indonesia
Expected DegreeB.Sc in InformaticsClass of 2027
Muhammad Nabil Hanif
M. Nabil HanifID // 2026

06 / Recognition

Honors and awards.

05 DOCUMENTED AWARDS

01 // 20252nd Place Winner

MetaMask Card Dev Cook-Off 2025

Track 4: Multichain USDC Payment System · Global Hackathon

Awarded 2nd place internationally with Freex, an automated cross-chain USDC settlement infrastructure for Web3 commerce. Contributed as Developer Relations, architecting technical documentation, developer community outreach, and protocol positioning.

02 // 2025APAC Representative

Google Solution Challenge 2025

Google Developer Groups x Asian Development Bank · Asia-Pacific

Represented GDSC UII across APAC with EcoSaver, an AI and OCR-driven mobile utility telemetry application optimizing sustainable consumption patterns under the UN Sustainable Development Goals.

Verified Record
03 // 20242nd Place Winner

Informatics Expo: Semester 2

Faculty of Industrial Technology, Universitas Islam Indonesia

Engineered desktop software architecture for regional travel and logistics orchestration, recognized for UX precision and clean modular software architecture.

Verified Record
04 // 2024Finalist and Exhibitor

Informatics Expo: Semester 3

Public Innovation Exhibition, Pakuwon Mall Jogja

Selected for public exhibition for developing an alumni employment analytics and professional networking platform bridging university talent with industry partners.

Verified Record
05 // 2023Finalist

Informatics Expo: Semester 1

Faculty of Industrial Technology, Universitas Islam Indonesia

Prototyped a rapid emergency healthcare dispatch application coordinating first-response triage and medical transport.

Verified Record

07 / Contact

Get in touch.

Interested in discussing a software project, retrieval architectures, or technical consulting? Feel free to reach out directly.

01 / Source CodeGitHub (848ill)02 / SocialInstagram (@bilhanif)03 / Inbox848.nabilhanif@gmail.com
Status & Availability
Open for Client Projects, Research, & Technical Consulting