Forward Deployed AI Engineer
Built production-grade AI and software for banking, fintech and enterprise.
I am open to freelance or part-time work in software and AI engineering. kurniadiahmadwijaya@gmail.com
~2 s to ~0.2 s
Knowledge-lookup interruption latency, controlled A/B (n=4 calls per arm)
ARIA, MII's voice AI contact center: ~2 s to ~0.2 s interruption latency in a controlled A/B (n=4 calls per arm), grounded answers and human handover.
8 engines
Scanning engines per pull request in a 100+ repository organization
Pris.AI, MII's AI code security platform: eight scanners per pull request, readable AI triage and a fix agent that opens PRs with validation and guardrails.
-23%
Reduction in prompt tokens; model calls per turn fell from 3 to 2
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.
MII Fraud Detection: OCR and two-pass LLM procurement review, with offline development-set evaluation. The hybrid engine remains in shadow mode only.
90.9% to 36.4%
False-positive rate on noisy clean packets, offline development set
eVendor, Metrodata's vendor portal for purchase orders, delivery acceptance, invoices, tax checks and approvals, built by one engineer in six weeks.
6 weeks
First commit to deployed portal, as the sole engineer
The MII AI Dashboard Template, a shared starting point for client AI apps, combining the web tier, agents, persistence, sign-in and infrastructure.
10+
Client projects that adopted the template
I am a Forward Deployed AI Engineer based in Jakarta, with 4+ years of experience building production systems for digital banking, fintech and enterprise SaaS. Since February 2026, I have led the delivery of Azure-native AI platforms at Metrodata for 12+ enterprise clients in sectors such as banking, automotive, mining, food manufacturing and consumer health. The work includes multi-agent RAG assistants, realtime voice agents, document AI and AI code security. Before that, I built backends and full-stack products for e-commerce, travel, SaaS and digital banking, including roles at MNC, Jatis and Jenius. I make LLM systems reliable in production through evaluation, guardrails and observability.
My employment, freelance work, education, research and awards, organized by year.
Oct 2026 to Present
LILT AI (United States, remote)
I design and build Indonesian-language tasks for Terminal-Bench, an agentic coding benchmark used to evaluate frontier AI models (OpenAI GPT, Google Gemini, Anthropic Claude) on real-world terminal and programming work. The tasks test whether AI agents handle Indonesian language and cultural conventions in realistic workplace data: government records, family registration, finance and procurement.
Feb 2026 to Present
Metrodata (PT Mitra Integrasi Informatika)
In the Microsoft AI practice, I lead Azure-native AI projects from architecture to production for 12+ enterprise clients in banking, automotive, mining, food manufacturing and consumer health. I also build the practice's own products, including a reusable full-stack AI template adopted across 10+ client projects. I write the specs, direct AI coding agents and review their work before release.
Nov 2025 to Apr 2026
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.
Nov 2025 to Dec 2025
TEMMPAT
I led the v2 backend refactor for a premium event ticket marketplace, separating auction and resale workflows into a modular engine.
Microsoft Agent Framework
LangGraph
LangChainAzure OpenAI (incl. Realtime)Prompt Shields
Hugging Face
pandas
NumPy
Matplotlib
Seaborn
Plotly
Streamlit
Jupyter
scikit-learn
PyTorchHugging Face Transformers
Keras
Azure AI Foundry
Azure AI SearchCosmos DBDocument Intelligence
Azure Container AppsAzure App Service
Azure Functions
Service BusKey VaultManaged Identity
Microsoft Fabric
Azure MonitorPythonTypeScript
JavaScript
Go
Java
PHPFastAPI
Spring Boot
Express.js
Gin
Echo
Laravel
CodeIgniter
Flask
Django
GraphQLNext.js
React
Vue
Nuxt
Tailwind CSSPostgreSQL
MySQL
MongoDB
Firebase
Redis
Kafka
RabbitMQ
Elasticsearch
DockerBicepGitHub Actions
Jenkins
AWS
LiveKit
k6
SonarQube
SentryThe AI coding agents Kurniadi works with, shown live: what each one is doing right now, plus public GitHub activity. The same office stands in a wing of the 3D Lobby.
The agents at their desks, today's numbers and the latest public GitHub activity, refreshed while this card is on screen.
My writing, research publications, published models and teaching materials.
Nov 2, 2026
First AI benchmark job: ten Indonesian tasks in ten days, half merged, most rated very hard, the top band. Over 300 ideas killed. The method, not the tasks.
Oct 27, 2026
ARIA moved from ACS to self-hosted LiveKit. A PoC drill measured 425 ms median hand-back over 20 rounds; no matched ACS-to-LiveKit speedup was measured.
Oct 24, 2026
An audit showed a repository's code could run inside the API holding production secrets. Same-day stopgap, then isolated workers and a 9-minute outage.
