Workflow Automation & Solution Engineering
Designing, building and deploying automation solutions across commercial training, higher education and third-sector environments.
PL-200 Power Platform Functional Consultant (Cert Ready) · Power Automate · Azure Logic Apps · Zapier · API integrations · SharePoint · Moodle · Arlo TMS
Overview
Automation work spans the full lifecycle: scope the problem → design the workflow → select tooling → build and test → deploy with training → evaluate. Projects have ranged from 24-step multi-branch admin flows to API integrations between enterprise platforms.
The guiding principle: the best automation is the one that solves the actual problem, can be maintained by others, and does not create new single points of failure. This means choosing platforms carefully, documenting thoroughly, and building in human oversight where it matters.
Case Study: IKON Training Digital Transformation
Designing and deploying a fully integrated Training Management System - Oct 2022 to Apr 2023
Challenge: IKON Training had no integrated digital infrastructure for bookings, learner registration, event communications, or post-course reporting - all administration was manual. The brief: deliver a fully integrated TMS in six months - sole internal technical lead, no external developers or contractors, go-live April 2023.
Outcome: A live, fully automated TMS integrating Arlo, SurveyMonkey, Power Automate, Moodle and SharePoint - with a Teams-based approval workflow and custom API integrations. All production flows migrated from Azure Logic Apps to Power Automate before handover, with full documentation for the team left behind.
Digital Development Strategy
The ADDIE-structured DDS - gave all stakeholders a clear picture of each phase and deliverables.
Course Communication Map
When communications sent, to whom, from which platform - for stakeholders without system access.
The IKON Training online delivery studio - built to support fully remote course delivery alongside the new TMS.
What the Go-Live Delivered
A fully automated, integrated system: delegates booked via Arlo, received automated pre-course communications, were provisioned into Moodle on attendance, and triggered post-course feedback via SurveyMonkey - all without manual intervention by the admin team. Trainers confirmed attendance via mobile app from any venue; commissioners received consolidated reports automatically.
"We have seen many benefits of using IKON Training's new booking system. It saves me so much time: I no longer have to forward emails to each of our delegates - they get all the information they need directly from IKON, up to date, accurate, and with notifications if things change… I get all the registration details, event details, feedback, and more sent to me automatically so I can monitor the quality of the training."
Analyse
Stakeholder interviews, surveys and workshops with trainers, administrators and the Operations Director. Reviewed prior documentation from an external development company that had attempted a bespoke TMS - their failures shaped the strategy away from a fully in-house build. Also drafted the organisation's first data privacy policy at this stage, with GDPR compliance built into the database schema from the start.
Design
Designed the data architecture, integration pathways, and automation workflows. After evaluating TMS options against cost-per-learner, scalability, and integration capability, Arlo was selected as the core platform.
TMS Software Comparison
Stakeholder-facing TMS comparison - features, costs and integration potential.
Database Design
Custom database schema designed with GDPR compliance at schema level.
Develop: From Zapier to Power Automate
Initially built as two 24-step Zapier automations. Zapier's one-month timeout limit meant events booked more than a month ahead would not trigger - a critical failure for a forward-looking events calendar. After tracking this systematically over six weeks, I transitioned to Power Automate: no time limits, native SharePoint integration, Teams-based approvals, and included in existing M365 licensing.
Azure Logic Apps was also evaluated - faster to develop, better error messages - but its unpredictable per-execution pricing made it unsuitable for high-frequency operational flows. All production automations were consolidated in Power Automate before handover.
Tracking Zapier Reliability
The evidence base that justified the move to Power Automate.
API Integrations
Standard connectors couldn't reach all of IKON's requirements. Using Arlo's developer documentation, I built custom Power Automate flows via HTTP connector and JSON parser to query the API directly. I also integrated the Moodle API - automating learner enrolment so that when a delegate was marked 'attended' in Arlo, they were automatically granted access to post-course resources in Moodle.
Implement & Evaluate
Go-live involved tailored training for trainers, administrators and stakeholders, supplemented by SharePoint guides, instructional videos and drop-in clinics. Post-launch evaluation identified improvements: reducing email reminder frequency, simplifying commissioner amendment, and enabling trainer synopses before leaving a venue.
Complex Branching in Power Automate
Conditional logic to handle variable Arlo API output - not achievable in Zapier's no-code environment.
Platform Comparison
Zapier
Good for quick prototyping. Fails at scale: one-month timeout kills forward-booked events, can't write to SharePoint, limited branching, rising per-automation costs.
Power Automate ✓
Right long-term choice for M365 environments. Included in existing licensing, no time limits, native SharePoint, complex branching, Teams approvals. Now the primary platform for all flows.
Azure Logic Apps
Faster to prototype, better debugging. Per-execution pricing is unpredictable at volume - unsuitable as the operational backbone. Best for high-complexity sub-flows or prototype builds.
Teaching Automation: Supporting UCL Staff to Build Their Own Flows
A recurrence-triggered flow: pulls from an external HTTP source, checks SharePoint, then creates or updates. This is the style of flow taught to UCL staff.
Beyond building automations directly, a significant part of the UCL role involves helping professional and academic staff build their own Power Automate capabilities - reducing dependency on a single developer and building institutional capacity. Sessions move from simple single-step notifications to multi-step flows with conditions and approvals, emphasising the logic of automation (triggers, conditions, actions) rather than just following a script.
For the data and reporting side of this work, see Data, Analytics & Reporting. For the AI integration strand, see AI for Learning & Assessment.