What I offer
Services
Targeting Singapore remote full time. Also open to contracts, freelance, and part time work when the impact is clear. Based in Jakarta, GMT+7, same working day as Singapore.
#01Remote · open to discuss
AI Engineering & LLM Integration
Agentic LLM workflows, search with retrieval augmentation, and systems that call tools, with orchestration across multiple models and full observability tracing.
What you get
- Agent loops with tool use and human escalation paths
- Hybrid RAG / search over your knowledge base
- Observability traces (LangSmith / Langfuse) so failures are debuggable
- Guardrails, eval hooks, and operator alerts where needed
Good fit when
- You need AI that ships in production, not a demo notebook
- Support, ops, or internal tools where latency and trust matter
- You want routing across multiple models with clear cost/latency tradeoffs
Outcome signal: Example: Incident Brain cut root cause diagnosis from ~120 minutes to under 10 via hybrid RAG + Slack escalation.
GoKotlinPythonPostgreSQLpgvectorLangSmith
#02Remote · open to discuss
Backend & API Engineering
Scalable REST APIs, microservices, payment integrations, and reliable backend systems using Kotlin, Java Spring Boot, Go, or Node.js.
What you get
- Service design, APIs, and data models that match the domain
- Payment / webhook flows with idempotency and structured logging
- Messaging (RabbitMQ / queues) and cache layers where they earn their keep
- Hardening for production: retries, alerts, and deployable runbooks
Good fit when
- Clinic, fintech, or transactional systems with high throughput
- You are splitting a monolith or stabilizing a critical path
- Payments, inventory, or loyalty logic that cannot silently corrupt
Outcome signal: Example: clinic POS + BRI/Xendit payments across 100+ branches, ~200k+ monthly transactions.
KotlinJava Spring BootGoNode.jsPostgreSQLRedisRabbitMQAWS
#03Remote · open to discuss
Full-Stack Web Development
End to end web products built with Next.js, TypeScript, and Tailwind CSS, from design handoff to production deployment.
What you get
- Product UI with clear information architecture
- Auth, dashboards, and integrations wired to a real backend
- Payments, file storage, and analytics when the product needs them
- Deployed environments your team can operate
Good fit when
- SaaS MVPs and internal tools that need to look intentional
- You have Figma (or rough wires) and need a fast path to users
- Landing + product surfaces that should convert, not just exist
Outcome signal: Example: Acaara, a digital invitation SaaS with templates, guest lists, and Midtrans payments in production.
Next.jsTypeScriptReactTailwind CSSPostgreSQLR2
#04Remote · open to discuss
System Analysis & Architecture
Requirements gathering, UML diagrams, data modelling, and technical documentation that bridges business needs with engineering solutions.
What you get
- Stakeholder workshops turned into approved specs
- Physical data models, DDL guidance, and integration maps
- Figma / flow mockups that engineers can build from
- Test scenarios that prove the design before big build spend
Good fit when
- Enterprise programs that span many modules and stakeholders
- Ambiguity is burning sprint capacity
- You need a translator between business language and engineering
Outcome signal: Example: PGN CRM & Billing: Account Management under Rp 40T+ annual revenue, 100% FSD sign off, zero rework cycles.
UMLFigmaSQLJava Spring BootOracle