Senior Infrastructure Engineer

Why join us?

Mitti by SafetyCulture is a global tech company, just not the kind you're picturing. Our almost 1,000-person team builds tools that make work better for the three billion frontline workers who keep the world moving. We've got big tech scale, without the sign-off layers or corporate theatre. This next chapter is about scaling smart, powered by AI, a clear vision and real ownership.

Big tech impact, without the big tech ick.

Role Purpose

This role exists to build Mitti's hybrid network, connecting our existing Amazon Web Services (AWS) and Google Cloud Platform (GCP) environments to a new physical colocation (colo) facility, and connecting that facility to the internet through our own multi-homed network. It's a greenfield build that starts with research and development (R&D) workloads, so we can prove the foundation before production follows. It's also a software engineering role: the automation built here has to scale from one site to a global footprint, so network and hardware operations are treated as a software problem from day one.

This is the first hire into a new team. You'll be primarily a specialist across networking and a generalist across infrastructure rather than a deep network only specialist, and this is not a people leadership role.

Not in scope: application-level reliability and service level objectives, sole ownership of the on-premises Kubernetes platform, and corporate IT. These are owned or shared with other teams.

Key Responsibilities

• Design, build and operate Mitti's multi-homed internet edge using Border Gateway Protocol (BGP) with multiple carriers, so traffic moves predictably and losing a path is a non-event.

• Design and run the data centre fabric: an Ethernet Virtual Private Network (EVPN) control plane over Virtual Extensible LAN (VXLAN) overlays with equal-cost multi-path (ECMP) forwarding, so the network scales horizontally and predictably.

• Connect AWS and GCP to the physical network with a routed, secure path (interconnect, transit design, Virtual Private Cloud integration, Domain Name System, load balancing and IP address management), defined in code and reproducible. Use it to prepare for a production integration with AWS & GCP.

• Contribute to enabling Mitti's own routed Internet Protocol (IP) address space, including address and Autonomous System Number (ASN) allocation through the relevant Regional Internet Registry, announcement, peering and route origin validation, and migrate traffic without disruption.

• Build the automation layer for network and hardware operations (provisioning, configuration, validation and change management as code), starting at one site but designed for a global footprint.

• Contribute to standing up on-premises Kubernetes on physical hardware with the Cloud Infrastructure Team, owning the network and hardware layer it runs on.

• Contribute to the colo environment build: hardware selection, cabling standards, out-of-band management, hardware lifecycle and supplier relationships. The physical build is straightforward and infrequent, so the emphasis is on the network.

• Make the estate observable and self-service with the Observability team, and write the designs, standards and runbooks that the team that follows will need.

Required Skills & Experience

Technical Skills

• Production experience running a multi-homed BGP network, including peering and sessions with providers, traffic engineering and path selection, and diagnosing issues such as link flapping.

• Strong IP networking fundamentals and packet-level debugging (Wireshark or equivalent).

• Designing and troubleshooting an EVPN/VXLAN overlay fabric with ECMP multipath forwarding.

• Strong Linux system administration, including comfort treating network devices as Linux systems. Experience with open network operating systems such as SONiC is a strong advantage.

• Working knowledge of Kubernetes, especially networking (Container Network Interface plugins, services, ingress and load balancer integration) in self-managed or bare-metal environments. We'd welcome someone whose strength is Kubernetes and who is keen to go deeper into networking.

• The ability to write working Python or Go to automate network and infrastructure tasks. Experience with Terraform and Ansible is a plus. Our backend teams work in Go, and the Cloud Infrastructure Team's Kubernetes work leans on it too.

• Experience building or operating cloud network architecture in AWS or GCP (VPC design, transit routing, private connectivity, VPNs).

• Hands-on familiarity with physical infrastructure (racking, cabling, power, out-of-band management through Baseboard Management Controllers). Experience with enterprise or data centre network vendors (Arista, Juniper, Cisco or similar) is helpful, and we're open to people willing to learn.

• A plus, not a prerequisite: public internet routing (Regional Internet Registry allocation, Resource Public Key Infrastructure and route origin validation) and kernel-bypass networking (Vector Packet Processing, Data Plane Development Kit).

Behavioural Skills

• Loves a blank sheet. Gets energy from building something from scratch rather than inheriting it, and makes decisions with incomplete information, revisiting them as evidence comes in.

• Self-directed. Works out what needs to happen next, drives it with minimal direction, and defers decisions until the evidence is in.

• Experiments properly. Builds a virtual lab, forms a hypothesis, tests it, documents the result, and treats cost as an engineering input.

• Scrappy and cost-conscious. Finds practical, economical ways to get things done rather than waiting for a full toolset, which suits a build where we start small and expand.

• Builds with others, not at them. Works with the engineers who'll use the automation, takes code review seriously in both directions, and treats the Cloud Infrastructure Team as a customer.

• Receptive to feedback and open about mistakes. Mistakes during the R&D phase are expected and safe to make. You'll own them, learn from them and share what you learned.

• Knows when to pull in help. Brings in specialists in Cloud Infrastructure, Observability, Security or backend engineering rather than pushing on alone.

AI Skills

• Uses AI tools every day for infrastructure code, automation and runbooks, including bootstrapping and validating device configurations, with sound judgement about when not to trust the output in a change-controlled network.

• Uses AI to speed up research and to investigate network incidents, for example correlating logs, events and observability signals, and builds small AI-assisted automations (alert enrichment, inventory parsing, first-draft configurations).

• Minimum expectation: a fluent, everyday user of AI tooling for engineering work.

Success Looks Like

• Within about six months, a first network is standing and has been reviewed and re-architected where needed. Mitti has a multi-homed internet edge with multiple carriers, running BGP.

• AWS and GCP are connected to the live colo facility over a routed path that is designed, documented and reproducible from code.

• The first colo site is running R&D workloads on an EVPN/VXLAN/ECMP fabric that is documented, instrumented and tested for failure behaviour.

• On-premises Kubernetes is running R&D workloads for the Cloud Infrastructure Team, and the network and hardware layer is solid enough that no one is worrying about it.

• Network and hardware operations run through automation that other engineers use and extend, and the design holds up when a second site is added. This gives leadership the evidence it needs to decide on bringing production workloads in.

Key Stakeholders

• Cloud Infrastructure Team (primary internal customer and on-premises Kubernetes partner)

• Engineering Manager, Platform (direct manager)

• Engineering Director (mentor and stakeholder for the colo and hardware scope)

…