AI-Enabled Automation Programme

Compounding productivity gains across back-office operations.

Client

Yiğitoğlu

Role

Digital Transformation & BI Manager

Period

2023 — 2026

Location

Istanbul, TR

Fig. 02 — Yiğitoğlu

/ A

Overview

Back-office functions were drowning in repetitive, low-judgement work — approvals, reconciliations, document extraction. Rather than run a one-off automation project, we built a programme: a portfolio of AI-enabled flows on Microsoft Power Platform with a shared operating model behind them.

Power AutomatePower AppsAI BuilderSharePointDataverse

/ B

Challenges

01Automation opportunities were scattered across every function; no shared prioritisation lens.
02Business owners wanted magic, not maintenance — few were prepared for a citizen-developer model.
03AI features carried real risk (extraction errors, model drift) that had to be governed, not ignored.
04Every quick win had to compound, not create shadow IT.

/ C

Approach

01Built an intake and scoring model to rank opportunities by hours saved, risk, and reusability.
02Established a Power Platform Centre of Excellence: naming, environments, ALM, and reviews.
03Rolled out AI Builder-powered flows for document extraction, approvals, and exception routing.
04Coached function leads into co-owning their automations, with the team as reviewers, not bottlenecks.

/ D

Outcomes

67%

Lift in operational productivity

40%

Cut in manual processing time on integrated flows

Governed platform, no shadow IT

"AI didn't replace anyone. It quietly removed the parts of the job nobody wanted, and gave the team back the hours where judgement actually matters."
— Behrouz Bagherzadeh, on the project

Next case · 03

Digital Workplace PWA

Yiğitoğlu

Read →