VP of Network Engineering

Job description

**1. ABOUT THE ROLE** The global marketplace for bandwidth. Share aggregates telco infrastructure into a single intelligence layer, giving ISPs, hyperscalers, and AI companies a seamless path to deliver connectivity into every home and business. We're hiring a VP of Network Engineering to lead the function that makes this real. This is the senior-most networking role at Share, reporting directly to the founders. You will own the entire technical estate (backbone, aggregation, access, peering, the systems they depend on, and the partner integrations they touch) and the team that builds and runs it. This is a role for someone who wants to create, not just operate. The mandate is to design the protocols, define the standards, and develop the reference architectures that an entire ecosystem of partner ISPs will inherit from us, and to leverage AI across the operational surface so the network behaves more like a system that thinks than a stack of boxes that's been configured. The expectation is still hands-on. You'll set technical direction, run a small high-leverage engineering team, and stay close enough to the metal that you can sit on a BGP debug call at 2am and not be the bottleneck. This is also a deeply partner-facing role: you'll absorb partner ISPs into our platform and push each one past whatever ceiling they walked in with. **2. WHAT YOU'LL OWN** **The network as a system** - Design the protocols, standards, and reference architectures that the rest of the ecosystem inherits from us. Partner ISPs across the continent will adopt the patterns you set (peering policy, hand-off contracts, service profiles, security posture) because conforming to Share's baseline is how they unlock the speeds we deliver. This is closer to standards work than to operations management: it ships as configuration, but it lives as doctrine. - Make the architectural calls that don't have a textbook answer yet: how a shared physical backbone is sliced across many partner ISPs at multi-tenant scale, how policy is expressed once and enforced everywhere, where the control plane goes next, what we retire, and what we keep. - Treat the network as a system that learns. Drive the use of AI in operations: anomaly detection across flow and streaming telemetry, predictive failure analysis on optical and routing layers, automated triage and runbook execution, traffic-engineering and peering decisions informed by models, and config-drift detection across partner hand-offs. You won't build the models from scratch, but you'll set the bar for where AI replaces human reaction time and where it doesn't. **The routing and peering core** - Own the strategy for our BGP edge: IP transit with Workonline, WIOCC, MTN/Bayobab, Afr-IX, and Skytrend, plus public and private peering at KIXP and LINX, and PNIs with the hyperscalers. - Own the policy posture: filter chains, route policy, communities, RPKI, and traffic-engineering logic across our border fabric and downstream. - Direct MPLS in the backbone (L3VPN and EVPN for partner isolation, services delivery, and multi-tenant scale) and decide where the control plane goes next (LDP vs. SR, address-family strategy). - Negotiate and establish new peering arrangements (content, CDN on-nets, cache fills, hyperscaler PNIs) and tune the mix for cost, performance, and resilience. **Aggregation, access, and the services plane** - Own the disaggregated, white-box switching strategy: modern NOS stacks on commodity silicon (SONiC, IP Infusion OcNOS, and similar). You should have a position on vendor lock-in, and you should have already acted on it elsewhere. - Direct L2/L3 aggregation (VLANs, MC-LAG/EVPN-VXLAN, QoS, clean L2 hand-offs to partner ISPs) and BNG strategy (PPPoE/IPoE termination, QoS, CGNAT where needed, CoA-driven session control that hooks into our AAA layer). - Own passive optical access (GPON, XGS-PON, OLT/ONU strategy, service profile design) and the fiber plant itself: ODF and splice planning, OTDR traces, loss budgets, and dark-fiber troubleshooting across our leased routes (KETRACO, KPC, Systel, and the ADC, iColo, PAIX metro ring). - Partner with the Systems team on the AAA stack (FreeRADIUS proxy architecture, per-partner containerized instances, Splynx integrations, CoA relay for payment-driven session control). They own that stack; you own the network-side of every integration, every session-control decision, and every end-to-end accounting flow. - Own the network-layer services the AAA plane depends on: NAT/CGNAT, DHCP, IPAM, DNS, IP/ASN planning, IPv6, and dual-stack rollout. **The team** - Hire, develop, and lead the network engineering team, including the junior and senior engineers we'll add as we scale. You set the bar, the culture, and the bench. - Build a team that writes things down. Push our source-of-truth (Nautobot) and config automation (Jinja-based generation) forward with the Systems and Dev teams, so the network is driven by data rather than tribal knowledge. - Run a team that's allergic to ticket-shuffling and bureaucracy: runbooks, post-mortems, automation, and a real culture of operational discipline. **Partner ISPs** - Lead the technical relationship with our partner ISPs. You'll sit on senior calls with their leadership and engineering teams, walk their networks, and own the technical commitments we make on both sides of the hand-off. - Translate between “this is what we need to deliver 100x speeds on your network” and “this is the concrete change on your OLT, your router, or your peering config tomorrow morning.” Every partner shows up with their own vendor mix and habits (MikroTik on RouterOS, Huawei on access, all the variations); the job is to bring each one to our baseline and push them past whatever ceiling they walked in with. - Own the technical credibility of Share with partners. When their senior engineer is on the call, you should be the person they want on the other end. **Uptime and the broader platform** - Own PoP and backbone reliability end-to-end: SLOs, incident response, post-mortems, and the culture around them. Catch degradations before partners do. - Sit at the leadership table as the voice of the network. Inform product decisions, capital allocation on capacity, peering strategy, and how we expand geographically. **3. WHO WE'RE LOOKING FOR** We care much more about depth, judgment, and ownership than about org-chart titles on a résumé. The right person is a builder who happens to have grown into leadership, not an executive who has drifted away from the work. If you've spent the last few years inside a large telco running PowerPoint reviews two layers removed from the CLI, this isn't your role. If you're still close enough to your network to spot a leaky origin from the symptoms before someone hands you a ticket, keep reading. **Must-have** - Senior leadership experience inside an ISP, carrier, or comparable network operator: running a team, owning architecture, and answering for uptime. Not enterprise networking, and not a pure-corporate role detached from operations. - A real track record of building or leading network engineering teams: hiring, leveling, setting technical direction, and being the person the team escalates to. - Deep BGP: route policy, communities, RPKI, route reflection, upstream/peer selection, traffic engineering, and the instincts to debug a leaky origin at 2am, not just direct the people who do. - MPLS in production at scale (L3VPN and EVPN), including control-plane details (LDP, SR, BGP address families) and opinions on where the architecture goes next. - Real operational experience with passive optical access (GPON or XGS-PON), OLT/ONU provisioning, and service modelling. If you haven't run PON, you've run an access network at comparable scale and understand the protocol stack. - Comfortable in disaggregated, white-box environments (SONiC, IP Infusion OcNOS, or similar). You don't need a vendor badge to operate a switch. - Hands-on with MikroTik (RouterOS) and Huawei (access or core). These are on our partners' side of the fence and you'll touch them constantly; you can sit on a partner's CLI and read it as fluently as your own. - Fluent in L2/L3 fundamentals: VLANs, STP family, LACP, OSPF/IS-IS, VRFs, NAT, VPNs, QoS. - Comfortable driving troubleshooting end-to-end (tcpdump, Wireshark, iperf, MTR, packet captures on both ends of a session) and able to lead a team to do the same. - Partner- and customer-facing maturity. You can sit on a call with a partner's engineering team and their leadership, read their network, explain what needs to change, and do it without condescension. This role is technical, but not back-office. - A bias for execution. You'd rather ship and iterate than circulate a strategy memo. Hierarchy and corporate theatre exhaust you. - Based in Kenya, or ready to relocate. You'll be close to our facilities in Mombasa and Nairobi and on calls with partners in the same time zone. **Nice-to-have** - Experience running or contributing to SONiC, FRR, or other open-source networking stacks. - Automation chops (Python, Go, or Bash), especially against NETCONF/gNMI, Nautobot, or Jinja-driven config pipelines. - IPv6 planning and dual-stack rollout experience at subscriber scale. - Internet Exchange operations experience: looking-glass and route-server behaviour, IRR/RPKI hygiene, peering negotiations. - Working familiarity with FreeRADIUS, CoA/Disconnect flows, and RADIUS-driven billing integrations. Enough to debug jointly with the Systems team, not to own the stack. - Exposure to applying ML or AI in network operations (anomaly detection on telemetry, predictive failure analysis, automated remediation, model-informed traffic engineering). We're not looking for a researcher; we're looking for someone who's already had to decide where this stuff helps and where it just adds noise. - Any PtP/PtMP wireless background is a plus for partner conversations, but it isn't the job: we are not a WISP. - Prior experience as a founding or early-stage network leader at a growth-stage operator. **4. WHAT WE OFFER** - The senior networking seat at the table while we build the technical backbone of Africa's next-generation internet. The decisions you make will shape how an entire ecosystem of ISP partners run their networks, and what the standards for that ecosystem look like. - Real ownership of strategy, architecture, and execution. No layer of middle management between you and the work. - Executive-level compensation: competitive salary and meaningful equity in a mission-driven, investor-backed company (US-incorporated; Kenya operating entity). - Private health and wellness benefits. We'll walk through these during the process. - A high-ownership environment with a steep but well-supported learning curve, and a team that writes things down.