Nestlé Nigeria PLC · Project write-up

Nestlé, project by project.

Nine pieces of work across the main factory, Site 2 and the distribution centre: what went wrong or needed doing, what I did, and how it ended. This is the redacted version.

Redacted version

Why parts of this page are blacked out

This is a live factory. The full write-up covers network layouts, device counts, cutover sequencing and how recovery and access are handled: details that map out how a working site is built and protected. I don't publish those. The shape of each project and how it ended are all here. The specifics are shared on request, one reader at a time.

The full version is shared on request.

Request the full write-up

21 Feb 2025Join as IT Systems Support
Jan – Feb 2026Cisco to Aruba migration
Apr 2026Main link failure
Apr – Jul 2026Acting IT Coordinator
18 Sep 2026CWAR all-hands
NowSoft handover, still carrying a lot

01 · Jan – Feb 2026

Cisco to Aruba, the whole factory

Replaced the factory's whole wireless and switching estate, vendor to vendor, without stopping production.

300+ access points
~50 switches and routers
Almost 20 major buildings

Context: The factory ran on Cisco: access points, switches, routers. All of it was being replaced with Aruba.
Problem: Cisco and Aruba can't be connected the way like-for-like replacements can, so there was no clean swap. Some equipment was incompatible with the new setup in both directions, which meant . And the factory couldn't stop.
What I did: I planned and configured it: redundancy first, then stages. I brought up the new network beside the old one and moved one section at a time, starting with and ending with . Every move had ready in case it went wrong, and the whole thing ran as a side operation so production never had to care.
Outcome: About five minutes of downtime per access point, ten when a switch or an AP failed or something unexpected happened. Done across January and February 2026.
Withheld: Cutover sequencePer-area planRollback approachDevice inventory

02 · April 2026

The main link failure

The factory's main internet link failed and production stayed up.

Under 10 minutes to fail over
No production downtime

Problem: The provider's side of the primary link failed. Production depends on that link.
What I did: I moved the factory onto the secondary link in under ten minutes. It has about a quarter of the primary's capacity, so I held back mobile Wi-Fi to keep the wired network, where the core processes run, at full speed. I made the call to , watched until the primary came back, then moved everything back .
Outcome: No downtime on the IT side or in production. The trade-off was real and deliberate: laptops on Wi-Fi were slower and people told me. There is a formal incident report, and it is part of the full version.
Withheld: Incident timelineFailover designProvider details

03 · Mid-April – early July 2026

Acting IT Coordinator

Ran IT for the whole site for three months with four interns, in a seat built for three people.

3 months
4 interns
20+ tickets a day

Context: The coordinator was promoted, the first hire for the job backed out, and the replacement started late. The seat stayed empty for about three months.
The ask: A managerial role listing five years of experience. I had just over one. I applied, I didn't get it, and I think that was reasonable. I had already shown I could do the work, so I did it.
Scope: The main factory (all departments, staff, infrastructure and network), Site 2 (two new plants, one starting up and one under construction), and the distribution centre, which had been mine since about my third month.
The interns: Four of them. I kept them busy and learning and kept it professional, managing the friction that comes with a team of peers. Over time they ran the day-to-day ticket queue while I coordinated and moved onto the heavier work.
Outcome: The site ran through the gap. In July the new coordinator started and we did a soft handover. I'm still carrying a lot while she learns.

04 · 18 September 2026

The CWAR regional all-hands

Coordinated a quarterly regional IT all-hands for nearly fifty people, and recovered it when the date turned out to be wrong.

Nearly 50 people
Quarterly, across the region

Context: A quarterly event that brings together IT from the region's sites and countries.
What I did: I coordinated it almost single-handedly. I split the work into streams for the interns: art, media, imagery and content in one, slides in another. Their ideas made it into the programme; most of the ideation and nearly all of the execution was mine.
Problem: A week before the event, we found the team on the slides had the wrong date: 22 October, not 18 September. The five weeks of runway they thought they had didn't exist.
The fix: The ideation was already done, so what remained was execution. I took it on myself and doubled the pace. Execution holds up under pressure. Ideation doesn't.
Outcome: It ran on 18 September. Leaders across the region recognised the programme and the effort behind it.

05 · Since about month three

The distribution centre

A remote site of 20 to 30 people that can't stop, run mostly by phone.

20–30 people
Handheld scanners
Label printers

Problem: Distance, users who aren't technical (so small problems get bigger), and printers. A label printer going down stops the building, and the pressure to fix it immediately is intense.
What I did: I ran it single-handedly alongside my main-site duties: remote calls, troubleshooting, teaching people the software, and raising SAP issues with the SAP team, since SAP support is mostly outsourced. For hardware faults I called or sent an intern.
Outcome: No real downtime. I'm available fast, and I sort it fast.
Withheld: Device inventoryVendor and escalation contactsSAP specifics

06 · Site 2 build-out

Site 2: the server room

Hands-on for the server room of the new site: servers, switches, fibre and endpoints.

Servers and VMs
Switching
Fibre

What I did: I was hands-on for the server room: the servers, the switches, the fibre runs and the endpoints. That covered and configuring each switch to our standard: .
Boundaries: I wasn't hands-on for the access point and cabling deployment, but I was close to it.
Withheld: Server and switch countsFibre and cabling layoutSwitch standards and VLAN design

07 · For more than a year

Onboarding and IT induction

Every new hire and every intern goes through me before they get access.

Every full-time hire
Every intern

Problem: Every new hire is a security risk until they understand the rules, and the people who join are not all equally careful.
What I did: I'm the entry point for everyone who joins. I run the IT induction: secure IT, cyber-security and phishing awareness, information handling and access management. I screen them before their account is activated by , then onboard them with their details and roles. I do this myself because of the stakes; the interns take over afterwards with the tools, mail and first passwords.
Outcome: The screening catches people who weren't paying attention or misunderstood something, before they have access.
Withheld: Induction materialScreening methodAccess policy specifics

08 · Ongoing

Recovery, monitoring and compliance

The checks that make sure recovery works before anyone needs it.

Recovery drills
Backups
Monitoring

Recovery drills: Periodically I , simulate and run restores, so we know recovery works before we need it.
Backups
Monitoring: across the network and the access layer.
Server room: are kept and maintained.
Dashboards: I don't build the Power BI dashboards for compliance and asset governance. I act on what they show.
Withheld: Backup schemeMonitoring stackCompliance records

09 · Ongoing

Everyday support

The second-line work that keeps the rest of the site moving.

Second-level support
Provisioning

The daily work: Second-level support for what the service desk can't resolve, tracked in the ticketing system. Provisioning, hardware assignment and asset tracking, workstation stability across admin and production, , and network readiness and endpoint installs for newly commissioned areas.
The plotter: The engineering plotter had been down for months. I got the OEM involved and had the parts replaced.
Withheld: Tooling and asset details

Request the full write-up Back to the Nestlé section