Skip to content
About

From Salesforce to data & automation — the same goal: decisions leaders can defend.

I am Alex Esposito, a Data & Analytics consultant at Accenture, based in Rome, Italy. I started in 2020 implementing Salesforce (CPQ, Service, Sales), then moved into analytics, data quality and automation.

What I'm working on now

  • Process mining with Fluxicon Disco on opportunity lifecycles — turning event logs into automations that cut idle time.
  • Refining a CRM data-quality scorecard that scales across regions without becoming yet another dashboard no one opens.
  • Experimenting with AI assistants over governed Power BI datasets — useful answers, not hallucinated ones.

The path here

  1. 2020

    Salesforce CPQ — first implementation

    Joined Accenture and landed on a CPQ rollout for an industrial manufacturer. Learned how quoting, product rules and approvals compound across the quote-to-cash chain.

  2. 2021

    Service & Sales Cloud rollouts

    Two back-to-back deliveries: Service Cloud for a utility (case routing, Omni-Channel, Knowledge) and Sales Cloud for a bank (pipeline, Lightning, Flow).

  3. 2022

    Pivot to data & analytics

    Most of my time shifted to the reporting layer over Salesforce. First Power BI cockpit from scratch — semantic model, DAX, Power Query, a governed dataset the leadership team actually used.

  4. 2023

    CRM data quality as a program

    Built a recurring data quality framework: rules in Salesforce, a Power BI scorecard, ownership per region. Data quality became a monthly ritual, not a clean-up project.

  5. 2024

    Process mining & automation

    Oil & Gas client: used Fluxicon Disco over Salesforce OpportunityHistory to surface rework and idle time, then shipped the automations — expiry management, escalations, reassignments, daily digest emails.

  6. 2025

    PowerApps staffing portal

    Collapsed four parallel Excel processes (time reporting, allocations, availability, chargeability) into one PowerApps + SharePoint Lists model for ~200 users — role-based views, hierarchy-driven permissions, Power Automate reminders.

How I think about the work

Decisions first, data second.

Before any modelling, I map the decisions the deliverable has to support. Everything else follows.

Explainable beats clever.

A score a team overrides is worthless. I choose approaches the business can defend in their own words.

End with ownership, not dependency.

Docs, tests, a runbook, and a training session. The team should own it on Monday.

Numbers with a named owner.

Every metric has a definition, a test, and a human who signs off on it. That is how trust compounds.