Kurniadi Ahmad Wijaya
Forward Deployed AI Engineer
AI and software built for the demands of banking, fintech and enterprise.
Technical Consultant, Software & AI · Metrodata (PT Mitra Integrasi Informatika)
Skip to content
Kurniadi Ahmad Wijaya

Kurniadi Ahmad Wijaya

Forward Deployed AI Engineer

AI and software built for the demands of banking, fintech and enterprise.

  • 12+enterprise clients
  • 5+industries served with AI
  • 7cloud and AI certifications

Need a software or AI engineer? Hire me. kurniadiahmadwijaya@gmail.com

Or start with

Selected work

All 40 projects
  • Realtime Voice AI Contact Center

    ~0.2 s from ~2 s

    Knowledge-lookup interruption latency, controlled A/B (n=4 calls per arm)

    Grounded realtime voice AI: ~2 s to ~0.2 s interruption latency in a controlled A/B (n=4 calls per arm), knowledge-based answers and human handover.

    Wing
    AI Engineering
    Period
    May 2026 to Nov 2026

    11 screenshots

  • AI Code Security Platform

    628 PRs security-scanned

    Pull requests scanned and triaged before merge, across 24 repositories

    MII's AI code security platform: seven scanners per pull request, offline rules with source credits, and fixes previewed and rescanned before human review.

    Wing
    AI Engineering
    Period
    Jun 2026 to Nov 2026

    6 screenshots

  • AI Solutions Portfolio and Assistant

    355 ms from Not run

    Prompt-injection screening of 15 to 20 retrieved chunks, which a silent API limit had been skipping

    The bilingual portfolio site of MII's Microsoft AI practice, in dark and light themes, with a multi-agent RAG assistant that answers with citations and a /dashboard Studio for content and analytics.

    Wing
    AI Engineering
    Period
    Jul 2026 to Sep 2026
    Visit live site (opens in a new tab)

    9 screenshots

More key projects

  • AI Fraud Document Review

    AI Engineering, May 2026

    AI document fraud review: OCR and two-pass LLM procurement review, with offline development-set evaluation. The hybrid engine remains in shadow mode only.

    • Next.js
    • FastAPI
    • Azure OpenAI
    • Azure AI Document Intelligence
    • Microsoft Agent Framework
    • Azure AI Content Safety
    • Azure Cosmos DB
    • Azure Blob Storage
    • +7

    36.4% from 90.9%

    False-positive rate on noisy clean packets, offline development set

  • Full-Stack AI Application Template

    AI Engineering, Mar 2026

    The MII AI Dashboard Template, a shared starting point for client AI apps, combining the web tier, agents, persistence, sign-in and infrastructure.

    • Next.js
    • FastAPI
    • Microsoft Agent Framework
    • Azure Cosmos DB
    • Microsoft Entra ID
    • Managed Identity
    • Microsoft Azure
    • GitHub Actions

    10+

    Client projects that adopted the template

In short

I am a Forward Deployed AI Engineer based in Jakarta, with 4+ years building production systems for digital banking, fintech and enterprise SaaS. Since February 2026, I have led Azure-native AI delivery at Metrodata for 12+ enterprise clients across banking, automotive, mining, food manufacturing and consumer health. My work spans multi-agent RAG, realtime voice, document AI and code security. Before that, I built backends and full-stack products at MNC, Jatis and Jenius. Getting an LLM response is only part of the job: I use evaluation to decide what to ship, guardrails to keep the system within its limits, and observability to understand production failures.

Current role
Technical Consultant, Software & AIMetrodata (PT Mitra Integrasi Informatika), since Feb 2026
Location
Jakarta, IndonesiaGMT+7
Open to
I am open to freelance and part-time software and AI engineering work.

At a glance

  • 90.9% to 36.4%Fraud screening false positives: 90.9% to 36.4% on noisy clean packets (offline development set)
  • 11.7 MB to 156 KBLighter national summary dashboard: 11.7 MB to 156 KB per load on real data, 39 requests to 5
  • 20Competition placings, including 7 first places (2018 to 2022)

Volume, agent-assisted and human-reviewed: 1,190+ PRs merged in 2026, 11,503 GitHub contributions in the last 365 days.

Recent work

My employment, freelance work, education, research and awards, organized by year.

Full journey
  • Oct 2026 to Present

    AI Benchmark Engineer, Native Language Specialist (Indonesian, freelance)

    LILT AI (United States, remote)

    I design Indonesian-language tasks for Terminal-Bench, which evaluates frontier AI models (OpenAI GPT, Google Gemini, Anthropic Claude) on terminal and programming work. The tasks ask whether agents can handle Indonesian language and cultural conventions in realistic workplace data, including government records, family registration, finance and procurement.

    Freelance

  • Feb 2026 to Present

    Technical Consultant, Software & AI

    Metrodata logoMetrodata (PT Mitra Integrasi Informatika)

    In Metrodata's Microsoft AI practice, I lead Azure-native AI delivery from architecture to production for 12+ enterprise clients across banking, automotive, mining, food manufacturing and consumer health. I also build the practice's own products and a full-stack AI template adopted across 10+ client projects. AI coding agents implement much of the code; I own the specs, review and release decisions.

    Employment

  • Nov 2025 to Apr 2026

    Backend Engineer (agency team)

    Naremax

    I developed backend modules for a digital products and training e-commerce platform with a 5+ person agency team, working in a TypeScript monorepo.

    Freelance

  • Nov 2025 to Dec 2025

    Backend Engineer (freelance)

    TEMMPAT

    I led the v2 backend refactor for a premium event ticket marketplace. I separated auction and resale workflows into a modular engine, with transactions and inventory locks protecting checkout.

    Freelance

Certifications

  • GH-600GitHub, Oct 2026
  • AI-103Microsoft, Sep 2026
  • AI-901Microsoft, May 2026
  • AZ-900Microsoft, May 2026
  • ETS2024
  • IARC x APTIKOMDec 2021

All 10 certifications

Skills wall

  • AI and LLM

    • Microsoft Agent Framework
    • LangGraph
    • LangChain
    • Azure OpenAI (incl. Realtime)
    • Prompt Shields
    • Hugging Face
    • RAG
    • Hybrid/Vector Search
    • LLM evaluation
  • Data science and ML

    • pandas
    • NumPy
    • Matplotlib
    • Seaborn
    • Plotly
    • Streamlit
    • Jupyter
    • scikit-learn
    • PyTorch
    • Hugging Face Transformers
    • Keras
    • Sastrawi
  • Azure

    • Azure AI Foundry
    • Azure AI Search
    • Cosmos DB
    • Document Intelligence
    • Azure Container Apps
    • Azure App Service
    • Azure Functions
    • Service Bus
    • Key Vault
    • Managed Identity
    • Microsoft Fabric
    • Azure Monitor
  • Languages

    • Python
    • TypeScript
    • JavaScript
    • Go
    • Java
    • PHP
  • Backend

    • FastAPI
    • Spring Boot
    • Express.js
    • Gin
    • Echo
    • Laravel
    • CodeIgniter
    • Flask
    • Django
    • GraphQL
  • Frontend

    • Next.js
    • React
    • Vue
    • Nuxt
    • Tailwind CSS
  • Data and messaging

    • PostgreSQL
    • MySQL
    • MongoDB
    • Firebase
    • Redis
    • Kafka
    • RabbitMQ
    • Elasticsearch
  • DevOps

    • Docker
    • Bicep
    • GitHub Actions
    • Jenkins
    • AWS
    • LiveKit
    • k6
    • SonarQube
    • Sentry

Live Agent Office

Live activity from the AI coding agents Kurniadi works with, alongside public GitHub updates. Visit the same office in a wing of the 3D Lobby.

Opening the office...

The agents at their desks, today's numbers and the latest public GitHub activity, refreshed while this card is on screen.

Writing

My writing, research publications, published models and teaching materials.

All writing
  • Nov 9, 2026

    Before It Looked Like Talent

    A memory from university, a friend who learned faster, and the slow years of practice that came before anyone called it talent.

    • personal
    • growth

    EN, ID

  • Nov 2, 2026

    Ten Very Hard Benchmark Tasks in Ten Days

    First AI benchmark job: ten Indonesian tasks in ten days, all ten merged in the end, five after a redesign. Over 300 ideas killed. The method I used.

    • ai-benchmarks
    • llm-evaluation
    • ai-coding-agents
    • multi-agent-workflow

    EN, ID

  • Oct 27, 2026

    Grounded Voice AI: From ACS to LiveKit

    The AI voice agent moved from ACS to self-hosted LiveKit. A PoC drill measured 425 ms median hand-back over 20 rounds; no matched ACS speedup was measured.

    • voice-ai
    • azure-openai-realtime
    • evaluation
    • case-study

    EN, ID

Need a software or AI engineer? Hire me.

kurniadiahmadwijaya@gmail.com