AI Platform Engineering Manager

Creativechaos; Islamabad; On-site; PKR 9,773,806 - 16,755,096; Full time

Job summary

Creativechaos is looking for a platform engineer to manage an internal platform and lead AI workflows within the company. This role requires building and maintaining infrastructures, coding, and team leadership. Ideal candidates will have 5-10 years of experience in development or DevOps, with a strong background in using AI tools and maintaining accurate documentation. You'll engage directly with various department heads to translate their needs into technical specs, driving innovative AI solutions in daily operations.

Qualifications

• 5-10 years of experience transitioning from development to DevOps or platform engineering.
• Proven experience managing small teams actively involved in the work.
• Excellent organizational discipline ensuring up-to-date documentation and runbooks.

Responsibilities

• Manage infrastructure, including Proxmox clusters and AWS workloads.
• Build and ship internal tools, dashboards, and AI workflows.
• Lead IT operations and maintain productivity across 150 people.

Skills

Experience with Proxmox clusters

AWS workloads

Code documentation

Team management

Communication across functions

AI tooling in workflow

Tools

Proxmox

AWS

AI DevOps tools

Job description

Job Summary

This is a greenfield platform function for Creative Chaos and its subsidiaries. You will build and run the internal platform (infrastructure, internal tools, IT operations, and the people who maintain them) that runs AI agents alongside human engineers as a deliberate operating model. You are also the architect and manager of every AI workflow, tool, and agent the ventures inside the group rely on day to day. Not a company that’s planning to “explore AI next year.” A group where AI tooling is already in the workflow, and you will shape how it evolves.

Responsibilities

What you’ll own

• Infrastructure. Proxmox clusters, AWS workloads, self-hosted services (SSO, wikis, observability, internal apps). You decide the architecture, you ship the IaC, you own the runbooks.
• Internal apps and agents. When a venture needs an internal tool, dashboard, ai agent, or workflow automation, your team scopes it, builds it, and ships it. You write code on this team, not just review it.
• AI workflows across the group. Every venture has its own AI needs. You partner with the lead of each, scope what they need, design the workflow, and ship it. You are the person every department head comes to when they want AI in their work.
• IT operations. Identity, endpoint management, access reviews, the unglamorous reliability work that keeps 150 people productive.
• The translation layer. Stakeholders bring you problems in plain English. You turn them into technical specs your team can ship.

You’ll lead three direct reports today: one each for IT, infrastructure, and tools/apps. The team grows from there based on what the ventures need.

How you’ll work

The operating model is AI-native. You will work alongside an AI DevOps agent that maintains runbooks, proposes architecture, and reviews infrastructure changes. Claude or Cursor are daily tools for you and your team, the same way an IDE is. Infrastructure is code by default. Documentation is written as work happens, not after. If the idea of pairing with an AI agent on architecture review is interesting rather than uncomfortable, you’ll fit the room.

• 5-10 years of experience. Where you started as a developer, moved into DevOps or platform engineering, and never stopped writing code. You’ve been doing this long enough to have opinions about what works.
• You’ve managed small teams (2-5 people) and were good at it, but the management didn’t pull you out of the work.
• You have excellent organizational discipline. Drift is your enemy. SOPs, runbooks, and documentation stay current because you keep them current, not because someone reminds you.
• You communicate clearly across functions. You can sit with the head of sales, the head of finance, or the head of operations, listen past the stated problem to the real one, and walk out with a working spec.
• You already use AI coding tools in your daily work, not because someone told you to, but because you tried them and they made you better.
• You’re comfortable owning ambiguity. Nobody hands you a roadmap. You build it.

You’re not a fit if

• You haven’t shipped code in years and prefer it that way.
• You’re a deep specialist (pure SRE, pure observability, pure security) who wants to go deeper into one layer.
• You need a fully scoped backlog before you can start.
• You let documentation rot. You think SOPs are someone else’s job.
• AI in your daily workflow feels like noise rather than amplification.