AI Workflow

We are a Singapore-based venture studio that invests, builds, and operates across multiple real-world sectors, including asset allocation and capital oversight, blockchain and compute infrastructure, racing operations, and cross-region construction and supply chain projects.

This role is not about building an AI product to pitch or sell. It is about embedding AI directly into high-stakes operational workflows so our teams work with AI as an active teammate — today, not in the future.

You will help standardize and automate core processes , turning messy, real-world problems into reusable workflows, internal tools, and AI agents that are actually adopted and relied upon. You will have direct access to decision-makers, frontline context, and operational environments across the US, China, Japan, Southeast Asia, and Europe.

This is a hands-on, experimental role focused on shipping practical systems that carry real responsibility in live environments. The position is based in Singapore.

Key Responsibilities

Workflow Modeling & Prototyping

1. You will decompose a business line into "input processing decision output," and formalize it into a workflow that can be handed off and reused. Typical domains include:

2. Investing & asset management: post-investment tracking, quant/arb exposure and capital flow, risk surfacing, monthly/quarterly reporting Energy / mining / data center ops: output, power draw, anomaly alerts, operational efficiency

3. Racing & sim racing ops: race prep checklists, parts/gear status, logistics and critical notes, post-race review

4. Construction / build / supply chain: milestone tracking, vendor status, time-critical approvals and blockers.

5. Your goal is not just documentation. You're productizing that workflow into a "virtual teammate":

6. Scheduled operational briefs

7. Live checklists that keep themselves updated
8. Automated risk/exception alerts

9. Consolidated status summaries that point to "who needs to do what next"

10. You'll assemble MVPs using modern AI infrastructure, for example:

11. Calling LLM APIs (including tool/function calling and action execution)

12. Building domain-specific conversational assistants with contextual memory
13. Adding retrieval-augmented generation (RAG) and vector search so the assistant actually understands our contracts, ops logs, site notes Orchestrating agent / tool chains so the model doesn't just "chat," it executes, logs, and pushes updates
14. Delivering this as something teammates can literally work with day to day (an internal assistant, a briefing bot, a risk broadcast channel)
15. The success metric is simple: does the team actually start using it.

Knowledge Capture & Structuring

• You'll deal with voice notes, chat logs, emails, on-site verbal updates, ops dashboards, handwritten notes — chaos. Your job is to turn that into:

• Queryable

• Maintainable
• Handover-ready
• Machine-readable structured knowledge

This means enforcing consistent fields, timestamps, ownership, and state definitions so that systems can reference it, report on it, and hand it off — without relying on "who remembers."

Automation & Efficiency

• You'll identify where humans are still patching gaps manually — and then remove that manual layer.

• You'll define a concrete success condition for each automation, e.g. "At 09:00 daily we get a briefing with current exposure, anomalies, and named owners for next actions."

• You'll deliver something that hits that standard, then iterate it forward from "prototype" into "infrastructure."

Collaboration & Resource Pull

• You're not an anonymous execution resource sitting inside a silo. You are expected to:

• Tell us what's actually highest priority right now

• Tell us what data, access, or people you need
• Take a messy pain point from A to a reusable solution at B, with minimal supervision

Education

Bachelor's Degree in Computer Science / Software Engineering, or equivalent self-taught technical capability.

Skills & Requirements

• Based in Singapore; able to collaborate in-person at high bandwidth (fresh grads, valid work pass holders, PR welcome).
• Open to non-fixed travel: when required, you may fly from Singapore to on-site locations in China, Japan, Southeast Asia, the US, or Europe to understand real conditions on the ground.
• Bilingual in Chinese and English (written and spoken), able to extract and structure critical information in both languages.
• Hands-on engineering ability: data fetching, API integration, automation scripting, or lightweight internal tooling.
• Strong structuring and documentation skills: you can turn scattered, informal inputs into standardized assets instead of tribal knowledge.
• Strong self-management; comfortable in a goal-driven, not task-driven, environment.

Nice-to-haves

• You've built or maintained internal tools, automation scripts, dashboards, alert/notification bots, ops or risk briefs, or milestone trackers.
• You've played a product role, or personally drove a small internal system from idea to "people actually use this."
• You've standardized a process: you turned something informal and person-dependent into a template with scheduled outputs and clear ownership.
• Familiarity with any of: LLM APIs, vector retrieval / RAG, agent / tool-calling workflows, automation schedulers, internal chat/assistant interfaces.
• You're curious (or already interested) in at least one of:
• Investing, capital deployment, quant/arb strategy monitoring
• Mining, energy efficiency, data center operations
• Racing / sim racing preparation, execution, review
• Construction / build / supply chain / milestone control