Engineering Lead

About Ajentik Ajentik builds AI-powered future skills for healthcare: the systems and workflows that move information, support decisions, and improve care for patients in one of the most complex industries in the world. Our flagship, Elderwise , is an AI-assisted expert assessments platform preparing for prospective clinical validation across three leading Singapore hospitals. We are incubated at NUS Enterprise, alumni of Harvard's Health Systems Innovation Lab (Top 10), and participants in National GRIP Deeptech and raiSE EnergiSE SG60. The team is small and deliberately focused. Healthcare is highly regulated, high-stakes, and underserved by modern software. That is precisely where frontier AI has the most meaningful leverage. The role We're hiring our future Head of Engineering to own technical direction, build alongside the team, and help bring our products into the hands of clinicians, operators, and patients. You'll report to the CEO and partner closely with the CPO. This is a player-coach role at an early-stage company: you'll own architecture, write meaningful code, and grow the engineering function over time. You'll be senior enough to set direction and standards, and stay close enough to the code to keep your judgment sharp. We don't expect deep expertise across every area below. We're looking for strong fundamentals, sound judgment on what to build, buy, or hire for, and the ability to learn quickly in a regulated domain. What you'll own from day one

• CI/CD and release engineering. Fast, safe, auditable pipelines with policy and security gates appropriate to a regulated domain.
• DevOps and platform. Infrastructure-as-Code, environments, secrets, deployments, and on-call, with the goal of moving us from "works" to "operable by anyone on the team."
• Kubernetes across a hybrid-cloud footprint. Workload architecture, cluster posture, cost, and the discipline of knowing when to use it and when not to, including where workloads sit on public cloud versus partner-hosted or on-premise environments.
• Data model architecture. Schemas, entity relationships, multi-tenancy, audit trails, and migrations across our products, designed for clinical-grade integrity and long-lived evolution.
• Production reliability, observability, and incident response for AI systems operating under healthcare data obligations. SLOs, tracing, alerting, and a real post-incident practice.
• Architecture and technical direction across backend, web, mobile, and AI.
• Engineering quality. Code review, design review, and the team's overall engineering culture.
• Our AI engineering practice (agents, tool use over MCP surfaces, retrieval, evaluations, and model routing), shipped safely.
• Hands-on engineering. You'll continue to write meaningful code on a regular basis.

What you'll grow into (or hire for)

• A small team of strong engineers. You'll lead hiring, levelling, and mentorship.
• Compliance-aware infrastructure codified in IaC and well documented: ISO 27001, ISO 42001, SOC 2, HIPAA, and Singapore PDPA.
• A repeatable AI shipping pipeline: evaluation harness, prompt and model versioning, safe rollouts, and behavioural observability in production.
• Dedicated platform, security, and AI-engineering disciplines as the team grows.
• Healthcare-specific integration surface (FHIR R4, HL7, EHR connectivity), adopted where it solves a real problem.

Stack and tools We are pragmatic about technology; we choose what fits and change it when it stops fitting. Backend and AI services: Python and FastAPI. Web: SvelteKit, Gradio, and React. Mobile: Swift and Flutter. Data: PostgreSQL as the system of record, pgvector for embeddings, Supabase Edge Functions for low-latency request paths. Platform: Kubernetes for production workloads, OpenTofu for IaC, GitHub Actions evolving into a hardened release pipeline. AI: leading open-source models for clinical reasoning, embeddings, and voice; a 470M domain-specific decoder model; and MCP as the protocol for exposing tools to agents. We run a hybrid-cloud footprint : public cloud, where it earns its keep, partner-hosted or on-premise, where hospital data residency and integration constraints require it. The primary cloud for each surface is chosen based on data residency, partner constraints, and AI service availability. What success looks like First 90 days. You understand our stack, our customers, and our risks in depth. Production is measurably healthier than when you arrived. You pick the metric (incident rate, MTTR, change-failure rate) and we'll back it. You've made at least one well-reasoned call on technical direction that the founders agreed with, and one they pushed back on, both as productive conversations. Six months. Engineering operates on the standards you've set. Our deployment pipeline and AI shipping process (evaluations, rollouts, and behavioural observability) are becoming recognised reference points within healthcare AI. You've made your first engineering hire, or made a considered case for waiting. Twelve months. Engineering is a real function rather than a few people improvising. We ship faster and more safely than at month one, and the team can take on the company's next level of ambition without being dependent on any single person. You'll thrive here if you

• Have worked through genuine ambiguity at an early-stage company and have chosen to keep doing so.
• Have built and operated CI/CD and Kubernetes-based platforms that other engineers actually want to use.
• Treat data modelling and schema design as a first-class engineering discipline rather than something you do once and forget.
• Stay close to the AI frontier: read research, try new tools, and form defensible views on what is overhyped and what is underrated.
• Find healthcare's stakes motivating rather than daunting.
• Hold considered views on coding agents, MCP and tool-use design, evaluation methodology, and what good AI engineering looks like today.
• Treat compliance as engineering (codified, tested, and reviewed) rather than as paperwork.
• Care about developer experience as much as user experience, and have invested in both.
• Can move between architecture, code, and team leadership in the same week while maintaining quality across all three.

Nice to have

• Background in healthcare, life sciences, or another regulated industry (HIPAA, SOC 2, HITRUST, MOH or local equivalents).
• Hands-on Kubernetes in production (GKE, EKS, or AKS) and experience operating workloads across a hybrid-cloud footprint.
• Direct experience leading or contributing to SOC 2 , ISO 27001/42001 , and desirable knowledge of HIPAA compliance.
• Deep experience with PostgreSQL at scale: schema evolution, partitioning, performance, and operational maturity.
• Hands-on experience with MCP servers and clients, or comparable agent tool-use protocols.
• Open-source contributions in AI tooling, evaluation frameworks, developer tooling, or infrastructure.
• Public technical writing, talks, or side projects that demonstrate how you think.

How we work

• A small team with minimal bureaucracy.
• We move quickly when decisions need to be made, and communicate clearly when they take more time.
• "Disagree and commit" is our default; thoughtful objections are welcomed and engaged with seriously.
• Progress is measured by shipped value and safe outcomes rather than time spent at a desk.

Compensation and logistics

• Competitive base salary plus meaningful equity. Specific range shared in the first conversation.
• In-office full-time with flexible hours when work or life genuinely calls for it. Two locations: Temasek Shophouse, 28 Orchard Rd, or NUS i3, Hangar L3-G.
• Singapore work authorisation required, or sponsorable via Employment Pass for the right candidate.

How to apply Email careers@ajentik.ai with:

• A short note on why this role, specifically.

• Links to anything you've built, shipped, or written that you'd want us to see.

• Your earliest realistic start date.

No cover letter required. We read everything we receive and reply either way.