Team Lead

We are hiring a senior developer to run the technical build of iQLocus, an agentic location-intelligence product that helps large retailers decide where to open, close, and invest in stores. You will own the architecture end to end, coordinate a small group of specialist contractors working in parallel lanes, and ship a system in which teams of autonomous agents build, maintain, and reason over predictive models with minimal human hand-holding.

This role suits an engineer who has delivered enterprise software to demanding clients, thinks about systems as complex adaptive systems rather than fixed pipelines, and wants to build agent teams that learn and coordinate on their own.

About iQueue

iQueue.ai builds a MissionOps platform: shared infrastructure on which teams of AI agents are configured and deployed into specific business problems. The platform runs on CrewAI agents, a Redis event mesh, a six-layer Memory Fabric, and a Semantic Membrane Architecture that governs agents through membranes, guider agents, and a decision ledger. The same substrate already powers products in category management, food-service shrink, retail media, healthcare, and nonprofit operations.

iQLocus applies that platform to retail store networks. Its first engagement is with a leading Asian grocery retailer and covers two jobs:

• Growth planning: white-space analysis to find where new stores should go
• Portfolio optimisation: which existing stores to keep, close, relocate, or invest in, including validating store-level capex

What you'll do

You are the technical owner of the iQLocus build. iQLocus is organized into six layers, and you are accountable for how they fit together:

1. Data extraction and integration. Bring public location data, partner location datasets, and client store, sales, lease, and capex history into a governed digital twin, with the client's sensitive data kept in its own hosted store.
2. LLM fine-tuning and knowledge capture. Extract the client team's expertise and use it to fine-tune a model for client-specific agent reasoning.
3. Agent model building. Design deterministic agent workflows (skills, tools, workflows, agent roster) that build, back-test, and tune the predictive models, so humans set up and engineer rather than hand-fit.
4. Agent reasoning and trade-offs. Build the Director, memory, and counterfactual agents that turn model outputs into a growth strategy with rationale, risks, and alternatives.
5. Agent management and problem-solving. Wire agents into the platform's governance: membranes, guiders, the decision ledger, and human approval gates.
6. Event interception. Have agents watch streaming events (lease changes, competitor openings, regulation, sales shifts), triage them against the models, and recommend when the strategy should change.

Alongside the build, you will:

• Break the work into isolated swim lanes with clear contracts, and scope, brief, and review the contractors who own them (for example, one builds a generalized additive model, another configures agents)
• Set engineering standards for code review, testing, back-testing, observability, and deployment across a hybrid client-cloud setup
• Keep the fitted models cleanly decoupled from the agentic layer so the client can own, run, and update its models
• Track agent work and token usage, since pricing is built on agent work units
• Join client and partner working sessions to translate business questions into technical design
• Write the technical documentation, handoff material, and runbooks the client team needs to maintain its models

What you bring

Enterprise software development

• 8+ years building production software, including at least 3 leading the architecture of a system used by large enterprise clients
• Strong Python, plus solid API design, data modeling, and cloud deployment (containers, CI/CD, infrastructure as code)
• Experience with event-driven or streaming architectures, message brokers, and distributed state
• A track record of delivering against a client scope, milestones, and acceptance criteria, including security and data-privacy requirements for sensitive client data
• Experience leading engineers you don't directly manage: contractors, partners, or cross-functional teams

Complex adaptive systems

• Working familiarity with complex adaptive systems: emergence, feedback loops, self-organization, and non-linear behavior
• The judgment to know when a system should be deterministic and auditable, and when it should be allowed to adapt and learn
• Comfort reasoning about how many interacting agents produce system-level behavior, and how to observe, test, and govern it

Autonomous agent systems

• Hands-on experience building LLM-based agents in production, beyond a single prompt or chatbot: tool use, multi-step workflows, memory, and evaluation
• Genuine interest in multi-agent systems where teams of agents with different roles coordinate, delegate, and escalate to humans
• Familiarity with agent orchestration frameworks (CrewAI, LangGraph, AutoGen, or similar) and their trade-offs
• A practical view on agent evaluation, guardrails, and cost control (model routing, token budgets)

Working style

• You can explain a technical trade-off to a non-technical executive in two sentences
• You are comfortable in an early-stage company where you set the structure rather than inherit it

Nice to have

• Geospatial and location analytics: catchment and gravity models, geofencing, PostGIS, or GIS tooling
• Interpretable modeling: generalized additive models, gradient boosting, back-testing, and model documentation for client handover
• Retail, grocery, real estate, or capex decision analytics
• CrewAI and Redis specifically, or experience with event mesh and membrane-style governance patterns
• LLM fine-tuning, distillation, or knowledge extraction from subject-matter experts
• Agent reinforcement learning or self-improving agent loops
• Digital twins or knowledge graphs as a shared world model for agents
• Experience working across time zones with teams in Asia, Europe, and the US