Position Description
Founding Engineer, Agents & Experience
Public-facing title: AI Products & Agents
________________________________________
About WIITY.ai
WIITY builds AI-powered workforces for recruitment and talent acquisition, automating repetitive work across the hiring process to improve productivity, efficiency and capacity, while freeing people to focus on relationships, hiring outcomes and growth.
Our success depends on more than the technology we build. Every interaction with WIITY, from the platform to a customer workshop, should feel unmistakably WIITY. Customers need to trust the experience we create and feel confident adopting the new ways of working we introduce.
About Ella
Ella is our proprietary platform, built to transform the way companies hire talent.
Ella is where the AI workforce lives. It is a multi-tenant platform on AWS and Snowflake, designed so that one governed data foundation and one shared set of platform tools serve both conventional software and autonomous agents. Its scope is the whole recruitment workflow, not a slice of it: sourcing, screening and shortlisting, interviewing, offer, onboarding and compliance, timesheets, payroll and invoicing.
Almost every step in that workflow is repetitive, high volume and currently done by a person who would rather be doing something else. That is the opportunity. Ella turns those steps into digital workers that operate inside a recruiter's day, with the human kept in the loop where it matters and taken out of the loop where it does not.
This is one of two founding engineering hires. The architecture is a first draft and has not yet been agreed.
________________________________________
Role Purpose
The Founding Engineer, Agents & Experience owns everything the customer actually touches: the application shell and its modules, the services behind them, and the agent runtime that turns Ella from software into a digital workforce.
Where the platform role owns the foundation, this role owns what stands on it. It is accountable for whether a digital worker reliably does the job it was given, whether a recruiter can understand and trust what it did, and whether the whole thing feels like one coherent product rather than a set of features bolted together.
Together with the Founding Engineer, Platform & Data, this role is responsible for using AI to build agentic capability across the entire recruitment workflow, from sourcing through to payroll and invoicing. In practice that means shipping digital workers that source and shortlist candidates, chase and validate timesheets, run compliance checks, keep records current across systems, and carry a placement through to an invoice, then proving each of them works well enough to put in front of a customer.
This is a founding role, not an implementation role. The architecture in front of you is a first draft. You will have strong influence over what it becomes.
________________________________________
Role Allocation
- Around 30% agent runtime and the digital workforce: harness, orchestration, model routing, toolbox design, agent behaviour.
- Around 25% application experience: the single UI shell, the modules within it, and the cross-cutting capabilities that make it feel like one product.
- Around 20% module services and APIs: the deterministic path, the API gateway and backend for frontend.
- Around 15% governance and platform services at the application layer: identity, access, tenant configuration, entitlements and billing.
- Around 10% shared: agent evaluation and quality, production support rotation, architecture and technical direction set jointly with the team.
________________________________________
Core Responsibilities
1. Founding Engineer: Shape the Architecture
- Work with the interim CTO and the second founding engineer to agree the Ella architecture, rather than inherit it.
- Make and document the foundational calls in your area: agent runtime design, framework and model choices, frontend stack, API and service boundaries.
- Bring a strong point of view. Disagreement, argued well and settled quickly, is how we expect technical decisions to get made here.
- Take as much ownership as you can carry. This is a small team by design and the intent is that the technical team owns its own domain end to end.
2. Agent Runtime & the Digital Workforce
- Own the agent runtime and harness: containerised execution, state management, task execution, model routing and failure handling.
- Own how agents are assigned toolboxes and how least privilege is enforced at runtime, working with the platform engineer who owns the tools themselves.
- Build the digital workers that WIITY commercialises alongside human pod leads, including our own internal workforce (Ada, Ivy, Tess, Mae, Iris, Flo, June, Cleo and those that follow). Product direction for these sits with the COO. How they actually work sits with you.
- Extend agentic coverage across the recruitment workflow over time: sourcing and screening first, then interview coordination, onboarding and compliance, timesheets, payroll and invoicing. Each new digital worker should be cheaper and faster to stand up than the last, because the runtime and the toolboxes did the heavy lifting.
- Design agents around the recruiter's actual day. A digital worker that produces good output but breaks how a consultant works will not get adopted, and adoption is the product.
- Own the decision of when a task should run deterministically and when it needs an agent. Reaching for a model where an API call would do is a cost and reliability problem, not a sophistication signal.
- Own agent latency and cost per task, since our commercial model prices digital workers against human headcount.
3. Application Experience: One Shell, Many Modules
- Own the WIITY application: a single UI shell containing Analytics, Workflows, Recruitment, Finance and the modules that follow.
- Own the cross-cutting capabilities that make it feel like one product rather than several: global search, notifications, user preferences, help and support.
- Own the interaction design of agentic features. Making an autonomous system legible to a recruiter, so they can see what it did and why, is the hardest design problem in the product and it belongs to you.
- Build the surfaces that our regulatory position depends on: human-in-the-loop review gates, audit and explanation views, AI content marking and disclosure. Recruitment AI is high-risk territory and these are product features, not documentation.
- Work with the brand direction set by Head of Brand Experience so the product feels unmistakably WIITY.
4. Module Services & the Deterministic Path
- Own the module backend services and the APIs behind each module, along with the API gateway and backend for frontend.
- Consume the shared platform tools and gold data products rather than reaching around them. Where the abstraction is wrong, fix it with the platform engineer instead of routing past it.
- Own service reliability, performance and versioning as modules multiply.
- Rebuild existing bespoke customer solutions as modules on Ella, working alongside the platform engineer who owns the underlying data migration.
5. Governance & Platform Services at the Application Layer
- Own identity and access: SSO, OAuth, role-based access control and how tenant isolation is enforced in the application layer.
- Own tenant configuration, feature flags and per-customer branding.
- Own the policy engine: permissions, approvals and the guardrails that sit between a user or an agent and an action with consequences.
- Own billing and entitlements: subscriptions, usage metering and what each customer is entitled to.
- Make the build versus buy calls in this area with the interim CTO. Time spent rebuilding solved problems is time not spent on the product.
6. Agent Quality & Evaluation
- Lead how WIITY evaluates agent quality: evaluation harnesses, regression testing across prompt and model changes, and the release discipline that stops a good agent quietly becoming a bad one.
- Define what good looks like per digital worker, in terms a customer would recognise, and make it measurable.
- Own the feedback loop from production behaviour back into agent design, working with the platform engineer on the data and observability that make it possible.
7. AI-Native Ways of Working
- Use AI tooling aggressively to multiply your own output. We are hiring a small number of strong engineers precisely because we expect each of them to work at the scale of a team.
- Hold the review discipline that makes that safe: nothing reaches production without a human who understands it and can defend it.
- Contribute to how WIITY itself works with AI internally, since we are the first proof of our own product.
8. Run & Support
- Share the production support and on-call rotation with the rest of the technical team.
- Own incident response, root cause analysis and the reliability improvements that follow, within your domain.
- Maintain documentation to a standard that lets someone else carry the pager.
________________________________________
Role Boundaries
This role owns: the application shell and its modules, module backend services and APIs, the agent runtime and the digital workforce, toolbox assignment and runtime enforcement, identity and access, tenant configuration, entitlements and billing, the interaction design of agentic and compliance features, and agent evaluation and quality.
This role does not own: the Snowflake data platform, gold data products, ingestion and transformation, the shared platform tools and integration adapters, AWS infrastructure, data residency implementation; silver layer schema design, which sits with the COO; product direction and prioritisation; commercial agreements and pricing; company strategy and marketing.
This role shares: overall architecture direction with the interim CTO and the second founding engineer; the tool invocation and least-privilege model with the platform engineer; customer migration, where this role rebuilds the experience and the platform engineer moves the data; agent evaluation, where this role leads and the platform engineer provides the foundations; production support and on-call with the technical team. In a team this size, the boundaries above describe accountability, not the limit of what anyone is expected to pick up.
________________________________________
Key Measures of Success
- Agent reliability: digital workers completing their tasks correctly, consistently, and without silent degradation after changes.
- Product coherence: modules that feel like one platform, and agentic features a recruiter can understand and trust without training.
- Speed from specification to working capability, and how that trend moves as the platform matures.
- Existing customers successfully rebuilt as Ella modules, without loss of function.
- Cost and latency per task, tracked and trending in the right direction.
- Compliance and trust features shipped as product, ahead of the customer asking for them.
________________________________________
Experience & Capability
Required
- Production experience building with LLMs, including tool and function calling, orchestration and the practical work of making non-deterministic systems behave.
- Strong TypeScript and React, and the judgement to build an interface people can actually use.
- Backend and API design experience in a real product, not just glue code.
- Demonstrable use of agentic coding tools as a core part of how you work.
- Track record of owning something end to end in a small team or early-stage environment.
Highly regarded
- Multi-tenant SaaS experience, particularly identity, tenancy and entitlements.
- AWS Bedrock, MCP, RAG patterns and evaluation frameworks.
- Design sense strong enough to work without a product designer for a while.
- Recruitment, HR or workforce technology domain knowledge.
- Experience shipping into a regulated environment, or building auditability and human oversight into a product.
How you work
- You are opinionated about architecture and quick to move once a decision is made.
- You are sceptical of your own agents and want to prove they work rather than assume it.
- You are comfortable with ambiguity, a moving roadmap and a first draft architecture that you are expected to improve.
- You want the scope that comes with being early, including the parts that are not glamorous.
________________________________________
Working Arrangements
- Employment type: permanent, full time.
- Location: Manly, NSW. Three days per week onsite, with the balance flexible.
- Reporting line: reports to the CEO, who holds the CTO function on an interim basis until a permanent CTO is appointed.
- Participation in the production support and on-call rotation.
- Remuneration: base salary plus superannuation.
Pay: $140,000.00 – $170,000.00 per year
Work Location: In person
