Forward Deployed Engineer, APAC

Dayos, Inc.

Singapore

Hybrid

SGD 120,000 - 170,000

Full time

11 hours ago

Be an early applicant

Application generator

An application made for this job — a tailored resume and cover letter that speak straight to the posting.

Get past ATS filters

Benefits offered by this job

Equity

Unlimited PTO

Regional travel

Job summary

Dayos is assembling an early engineering team in Singapore to build AI agents on the Hero platform, integrating with Oracle, Workday, and SAP. You will work directly with customers, ship production agents, and improve productized Starter agents across real enterprise data.

You’ll collaborate with the engineering lead and product teams, travel across APAC as accounts mature, and help shape platform standards while growing into a senior engineer role.

Qualifications

Evidence of unusual performance and ability to ship real agent systems.
Strong Python and scripting capabilities with production-grade code.
Experience designing and deploying AI agents for enterprise customers.

Responsibilities

Build production-grade agents on Hero for real enterprise workflows.
Design and ship custom agents for Pro plan customers.
Own and harden Starter agents in production.

Skills

Python

TypeScript

React

LLMs

Prompt engineering

Tool calling

Docker

Git

Linux

Oracle exposure

Tools

Google ADK

Gemini

Claude

LangChain

LlamaIndex

Job description

Dayos is the AI-first version of the systems integrators and application management firms that have supported enterprise ERPs for decades. We're building this company in Singapore, for the long term.

Hero, our platform, doesn't just watch your systems — it logs in, does the work, and validates the output inside Oracle, Workday, and SAP. At a Singapore enterprise, Hero closed 9 of 10 support tickets that had been open for an average of three weeks, taking roughly 10 minutes per request.

Enterprises spend billions every year supporting Oracle, Workday, and SAP. Much of that work is still manual. We're turning more of it into software. Three years building Hero. Enterprise customers already in production. Now we're closing our seed round and building the Singapore team from the ground up.

— 02 / The role

In production, with a real customer at the end of every ticket

This is not a research role. It's not building internal tools no one uses. You'll build and customize AI agents on top of Hero, and every agent has a customer using it and an outcome measured against it.

Two sides of the work. For Pro plan customers, you'll build new agents customized to their specific workflows — these are the customers who signed a contractually committed 60% reduction in their ticket queue, and your code makes that commitment real. For Starter plan customers, you'll harden and improve the productized agents that ship out of the box. When one of them breaks against a real enterprise's data, you're the one who fixes it.

You'll spend real time inside customer environments, on calls with customer stakeholders, and looking at production data. You report directly to our founder.

You're not joining an established engineering machine. You're helping build one. You'll be among the first Forward Deployed Engineers on our team, and you'll shape the engineering standards, the deployment playbook, and the agent architecture that every hire after you will inherit.

This will be demanding. Real production systems break in ways you didn't plan for. Enterprise customers ask questions no textbook prepared you to answer. We'll give you unusual responsibility early, and we expect you to grow into it quickly.

— 03 / Who you'll work with

People who have actually built this before

Our founder led Finance Engineering at Robinhood through its IPO, overseeing the same enterprise stack Hero now operates across: Oracle, Workday, Coupa, and Anaplan. Before that he spent 20+ years in Oracle consulting, running enterprise implementations at Fortune 500 companies.

On the engineering side, you'll work alongside our AI Engineering Lead and our Head of AI Product Engineering, both of whom have shipped production agent systems and know where the sharp edges are. You'll spend your first year learning from people who have actually built this before.

— 04 / What you'll do

Build agents that do the work

01 Build agents on the Hero platform using Google ADK, Gemini, and Claude. Prompt engineering, tool design, reasoning pipelines, and evaluation
02 Design and ship custom agents for Pro plan customers against specific enterprise workflows: reconciliations, AP approvals, report development, configuration changes
03 Own the productized Starter agents in production. Debug, improve, and harden them against real enterprise data
04 Integrate with Oracle Fusion, Workday, and SAP APIs. When the API is missing, build the workaround
05 Work directly with customer IT, finance, procurement, and HR stakeholders. Listen, then translate what they need into agent behavior
06 Contribute to the core Hero platform. Reasoning improvements, observability, token budget controls, and change management primitives
07 Travel to customer sites across APAC as accounts mature

— 05 / Your first year

From shadowing to leading your own customer projects

Months 0–3

Learn Hero end to end, the Oracle/Workday/SAP domain, and our forward-deployed model. Shadow senior engineers on active deployments. Ship your first agent improvements under close review.

Months 3–6

Own bug investigation and agent improvements on live customer accounts. Start pairing with customer stakeholders on scoped features.

Months 6–12

Lead your own customer projects. Design and ship custom agents for Pro plan accounts. Own an area of the productized Starter agents.

12+ months

Senior engineer trajectory. Deeper technical ownership, mentorship of the next cohort, and a voice in how the platform evolves.

— 06 / What you'll learn

By the end of your first year

Design, build, and ship production-grade agent systems on top of modern LLM stacks
Reason about token economics, evaluation, and observability at a level most engineers won't touch for years
Work fluently across the Oracle, Workday, and SAP API surface — including where the vendors' documentation lies
Debug production agents against real enterprise data, at real enterprise scale
Translate between senior finance, IT, and HR stakeholders and the code that has to make their world work
Hold your own in customer-facing technical conversations at the executive level

Weekly technical reviews and direct coaching from the founder and our senior engineers.

— 07 / What you bring

Charisma, hustle, and code that ships

Charisma and hustle
You read a room fast, you're comfortable talking to customers, and you don't wait for someone to hand you the next thing to build.
Evidence of unusual performance
Top of your class, published research, hackathon wins, open-source contributions, shipped side projects other people use, competitive internships — anywhere the results were measurable and stood out.
Fluent in Python
Comfortable in TypeScript and React. You've built real applications, not just done coursework.
Hands-on with LLMs
Prompt engineering and tool calling. Familiarity with agent frameworks — Google ADK, LangChain, LlamaIndex, or equivalent.
RAG and vector databases
Familiarity with Pinecone, Weaviate, or similar.
Docker, git, and comfort in a Linux environment.
High agency
When you see a bug outside your ticket, a customer friction point, or an infrastructure bottleneck, you fix it without being asked.
Fluent with AI-assisted development
Cursor, Claude Code, Copilot, or your tool of choice. You use them daily and know how to get more out of them than most.
Bonus: Oracle, Workday, or SAP exposure
Coursework, an internship, or a side project all count.