Forward Deployed Engineer (KL-based)

About WhiteCoat

WhiteCoat is a Singapore-headquartered omnichannel provider of integrated health andwellness services that serves as the first and single touchpoint for all care needs in SoutheastAsia.

Since launching in 2018 WhiteCoats digital platform powers a wide range of services includingtele- and in-person consultations as well as medication fulfilment and diagnostic testingacross primary specialist and allied care. With a focus on the B2B space WhiteCoat has forged strategic partnerships with the regionsleading insurers corporates and care providers to provide accessible and affordable high-quality care to its users.

The Group currently has offices in Singapore Indonesia Malaysia and more information on WhiteCoat please visit .

What you will be doing

We are seeking a Forward Deployed Engineer who can sit close to real business and healthcare workflows clarify ambiguous requirements design the solution build working software produce tests generate QA evidence and maintain the AI/agent/KB systems that make delivery faster and safer.

This is a client-facing AI-native delivery role. You will embed with client and business teams convert ambiguous operational problems into agent-executable work and operate specialised AI agents as the default delivery workforce.

You are not joining to manually process an engineering backlog. You are joining to design govern and continuously improve an AI-native delivery system that takes client problems from discovery through adoption and measurable production outcomes. You remain accountable for every decision claim release and business result produced through that system.

On a day-to-day basis you will be responsible for

• Business and client discovery
  Work directly with clients and commercial product operations QA security DPO and platform stakeholders. Turn vague asks into precise agent-ready briefs covering the business objective affected users and systems workflows market rules data classifications expected value acceptance criteria risks blockers and decision owners.

• AI-led solution design
  Direct agents to map repositories systems APIs data flows permissions audit requirements exception paths monitoring and rollout and rollback controls. Decide whether the right answer is configuration a workflow change an integration a prototype a client-specific exception a production feature or a reusable platform capability.

• Agent-orchestrated delivery
  Decompose business outcomes into agent-owned workstreams assign the appropriate context tools permissions and acceptance criteria and manage dependencies and handoffs between agents. Reconcile conflicting outputs and intervene when risk novelty or agent limitations require human judgement.

• Evals QA and release evidence
  Define edge cases and pass/fail criteria before delivery. Require agents to execute functional negative regression API and contract market-specific permission privacy and healthcare-workflow tests. Suggested tests do not countonly executed tests with reproducible evidence support a release decision.

• Healthcare data and production risk
  Ensure agents operate within approved boundaries for PII PHI NRIC clinical claims insurer pharmacy and payment data. Recognise when data must not enter an AI system and when security or DPO escalation is mandatory.

• Client-facing commercial execution
  Explain options constraints risks and trade-offs clearly to technical and non-technical stakeholders. Challenge requirements that are vague unsafe low-value or unnecessarily bespoke. Keep delivery tied to business value and never invent approvals test results security clearance or production readiness.

• Agent and knowledge-system improvement
  Turn defects UAT failures unsafe assumptions and delivery friction into stronger prompts agent workflows golden tasks adversarial cases regression suites KB rules SOPs and release gates. Every engagement should improve the delivery systemnot merely complete the immediate request.

• Agent governance

  Define what each agent may access decide change test communicate or release autonomously. Maintain least-privilege access healthcare-data boundaries auditable intervention points and mandatory approval gates for security DPO financial client-facing and production-impacting actions. Agent actions must be traceable reversible and evidence-backed.

What success looks like

Success means operating as an AI-native business-facing Forward Deployed Engineer who turns ambiguous client and commercial problems into working measurable outcomes.

Own the path from client problem to measurable adoption: establish the baseline define the value hypothesis agree the success measure and decision owner validate the workflow with real users and determine whether to scale revise or stop. Support rollout operational enablement incident communication and post-release adoptionnot merely technical handover.

Our Benefits

• Make a Real Impact: Opportunity to contribute to a leading digital health companys rapid growth.

• Fast-paced Start-up Environment: Experience an environment where you get to own and make tangible impact without bureaucracy getting in the way of rapid decision-making.

• Great Team: Collaborate with intelligent friendly and supportive professionals from diverse backgrounds.

• Hands-on Learning & Growth: Gain hands-on experience in strategy partnerships operations and product innovation within a growing industry.

• Competitive Compensation & Benefits: Competitive compensation and performance-based bonus. Holistic health insurance for your peace of mind for both in-patient and out-patient coverage.

How to apply

If you believe you have what it takes for this role click Apply and join us on our journey to make a positive impact on the lives of people through innovative healthcare solutions!

What we are looking for

Required:

• AI-native by default. Advanced hands-on experience using AI agents as the primary delivery layer across intake repo mapping solution design implementation QA security review release and remediationnot merely as autocomplete.

• Agent orchestration expertise. Able to design and operate specialised agents for backend frontend QA security/DPO release and healthcare workflows such as eligibility claims pharmacy and payments.

• Strong systems and production judgement. Able to direct agents across unfamiliar repositories APIs databases integrations CI pipelines and logs while controlling scope and assessing production risk. This is not a conventional manual engineering role.

• Strong AI SDLC expertise. Experienced with prompt and version control agent observability rollout and rollback controls RBAC consent and incident management.

• Strong eval discipline. Able to build golden tasks adversarial and trap cases rubrics pass/fail gates regression suites and remediation or training loops.

• Strong knowledge-base discipline. Turns failures and validated learnings into durable rules checklists SOPs agent instructions and quality gates.

• Evidence-led QA. Defines edge cases upfront runs exact verification steps preserves outputs and produces reproducible evidence others can trust.

• Sound healthcare data judgement. Understands the boundaries around PII PHI NRIC clinical claims and insurer data including when AI use is prohibited or requires DPO or security escalation.

• Strong healthcare workflow reasoning. Can reason across payer rules eligibility benefits claims pharmacy fulfilment provider and TPA operations payment states exceptions consent and audit trails.

• Relevant healthcare domain capability. Able to operate effectively in healthtech insurtech payer/provider platforms claims pharmacy fulfilment benefits administration or telehealth.

• Effective in high-ambiguity environments. Challenges vague requirements surfaces cross-functional trade-offs and reaches the right answer through structured building testing and evaluation.

• Evidence-based communication. Clearly separates facts assumptions risks decisions and unresolved gapsand never invents approvals results evidence or production readiness.

This role suits builders who own outcomes end-to-end and may not be a fit for those who only write requirements or who treat AI tools as autocomplete without owning the output.