From Scattered Data to Smarter Business Decisions

Customer logo
Case study thumbnail
DTI.al is a technology training academy in Tirana. Its whole student journey, from the first Meta enquiry to the last instalment, runs through the CRM. The team had the data but not the answers, so we built a reporting and operations layer on top of the CRM they already used.

Founded in

Tirana, Albania

Became Customer

Website

About DTI

DTI.al is a technology training academy in Tirana. It runs professional courses in areas such as data analytics, software development, IT and digital marketing. Most students pay through instalment plans of up to six payments, and most of them first reach the academy through Meta ads, start a conversation, talk to a sales advisor and then sign a contract.

Every step of that journey, from the first enquiry to the last instalment, was already tracked in the CRM. The academy did not lack data. What it lacked was a reliable way to turn that data into answers the team could act on every day.

The questions nobody could answer quickly

Every week management asked the same things. How many students enrolled this month? How much money is overdue, and with whom? What has each salesperson earned in commission? What did the last season of advertising actually bring in? Each of these questions meant exports, spreadsheets and a few hours of someone's time, and two people working on the same question could easily arrive at two different numbers.

Planning suffered for the same reason. Without a clear view of past seasons, it was hard to decide how much to spend on ads for the next intake, how many classes to open, or whether the sales team could handle more demand.

Where the real problem was

Once we looked at the data closely, it became clear that the numbers were not just slow to get. Some of them were misleading.

  • The CRM marked a deal as won only when the final instalment was paid. A student who enrolled in October could appear as a sale months later, which shifted every monthly chart.

  • Each contract stored up to six instalments in eighteen separate custom fields. Receivables, collections and overdue amounts had to be rebuilt from those fields by hand.

  • Ad platforms reported conversations, not real enquiries. Only 47.6% of Meta conversations ever reached the CRM as a lead, so cost per lead in the ad manager looked far better than reality.

  • Older records imported from spreadsheets carried swapped days and months, shifted timestamps and duplicates.

  • Eight automations were already running on top of the CRM, so every change had to be deliberate, visible and reversible.

Our approach

We kept the CRM as the system of record. The sales team did not have to change how they work, and nothing was migrated. Instead, we sat down with the academy's owner and wrote the business rules down once: when a student counts as enrolled, which season an enrolment belongs to, how commission is earned and when a payment counts as late. The guiding rule was simple: never chart a field before knowing what it means for this business.

What the team works with today

The result is a set of 14 views that each answer a question somebody at the academy asks every week:

  • A day-by-day overview of enrolments, cash and pipeline, with short plain-language insights.

  • A contracts board with one row per contract and one box per instalment. Ticking a box records the payment date, amount and status directly in the CRM.

  • Commission reports showing what each salesperson has earned and what is still pending, based on money that has actually come in, with refunds held back for review.

  • Marketing reports that follow ad spend through conversations and leads all the way to enrolments, by course and intake.

  • Season planning that turns a budget per course into expected students, cash, classes and teachers, using conversion and collection rates from real enrolments.

  • A costed model of the sales day that shows how much incoming demand the advisors can realistically handle.

  • A separate B2B pipeline for company clients, with a calendar that tells the team who to call next.

  • A snapshot view that compares any two points in time for any metric.

Under the hood: a data platform, not just a dashboard

Behind the views sits a small but complete data platform. Every 15 minutes, Solverix OS takes a full snapshot of the CRM together with Meta ad data. This is a classic extract, load and transform (ELT) flow: data is copied as it is, then cleaned and modelled in a separate layer where the business rules live. Fourteen days of snapshots are kept, so a bad sync can never rewrite the past and the team can always compare today with any recent moment.

Because the rules are defined once in that model layer, a chart and its CSV export always agree. A full rebuild of the model takes about a second, and the whole system runs on a dedicated server for the academy.

On top of the model run 49 automated data-quality checks. They catch swapped dates, payment schedules that do not add up, statuses that contradict payments and duplicate leads. 44 of them can be fixed directly from the dashboard, and every fix is logged with the value it replaced so it can be undone.

Why not just Power BI?

A tool like Power BI or Microsoft Fabric would have given DTI charts on top of the same unreliable fields. The real work here was understanding what the data means, fixing it at the source and letting the team act on it, for example by recording a payment from the contracts board. That needs write-back to the CRM, role-based actions, an audit trail and an undo, which a standard BI report does not provide. A custom platform let us build exactly the workflows the academy runs on, and nothing it does not need.

Fixing the data at the source

A good part of the work happened in the CRM itself. We added a proper enrolment date to 311 contracts, so months, cohorts and sales cycles are now measured from the signature, not the final payment. 231 missing season codes were recovered by learning from similar records, matching 95% of the known labels, and duplicates and refund ratios were corrected. The CRM is now more accurate on its own, not only inside the reports.

Built for the team to keep growing
  • The owner and admins can adjust business rules, such as season definitions or campaign mappings, from settings rather than code.

  • Eight named accounts with role-based access for operations, B2B and data fixes replaced a single shared login.

  • New needs are handled as small, regular releases. The whole system went from first sync to the B2B calendar in production in about five weeks.

The result

Management now looks at one set of numbers for enrolments, collections, commissions and marketing, and those numbers match the CRM. Overdue payments are visible early and are followed up while they are still easy to collect. Recording a payment takes one tick. Marketing decisions rest on enquiries that actually reached the sales team rather than on platform estimates, and season planning starts from real history instead of guesswork.

Closing perspective

DTI did not need a new CRM. It needed its existing tools to give consistent, trustworthy answers and to support daily work. By agreeing on the business rules first and building a proper data layer underneath, the academy turned a system that only stored data into the place the whole team now runs the business from.

These guys are great! They really know their stuff. Solverix has completely changed how we track and analyze our business performance. They've got the know-how, the technology, and a real understanding of what businesses actually need. Highly recommend!

Senada Kiri

Head of Partnerships, DTI

On this page

Keep exploring how modern teams ship

Keep exploring how modern teams ship