# A project still at the discovery stage

> Discovery for applied AI in tax consultancy, presented honestly — an active project, with no invented impact figures.

URL: https://thenichesociety.ro/en/case-studies/rsm-romania

Not every project in our portfolio is finished. This one is active — a discovery conversation about how applied AI could be used in tax consultancy. **We're showing it as it is, without inventing a result that doesn't exist yet.**

## A tax consulting and audit firm, at the first conversation

The conversation started from a concrete interest in AI applied to tax consulting work — not an abstract curiosity, but a question about where AI actually helps consultants' day-to-day work.

## Where AI concretely helps at a tax & consulting firm

Tax consulting work involves a lot of document handling, constantly changing legislation, and a need for high precision — exactly the area where AI applied correctly (not generically) can cut time spent on repetitive tasks, while keeping human review on the decisions that matter.

## The FOX precedent, as a starting point

We used our previous AI training work (the FOX project) as a concrete reference point — not a theoretical pitch, but an already-delivered example of what AI training done right looks like for a team.

We're at the stage of identifying the concrete use cases that deserve a pilot — we haven't reached implementation at scale yet.

## From the initial conversation to scoping a pilot

- 01Initial conversation, starting from the FOX precedent (AI training already delivered).
- 02Exploring together the areas where applied AI concretely helps tax consultancy work.
- 03Identifying candidate use cases for a pilot.
- 04Current stage: selection and scoping for the pilot.

## Honesty over marketing

We could wait until we have a “result” to show. We'd rather be transparent about the real stage — for someone evaluating whether to work with us, the process matters just as much as the final result. This page will be updated as the project progresses, with verified facts, not projections.

## What we used

There's no platform technical stack yet — we're in the discovery stage, where the discussion is about possible use cases, not architecture. The technical reference used in the discussion is the training already delivered for FOX.

## Why we publish an unfinished project

The kind of initial conversation in this project resembles how the relationship with FOX started too — which gives us confidence that the following stages, if they come, will be just as solid. But we don't anticipate results that don't exist yet, just to have a more “complete” page.

## Frequently asked questions

### What's been delivered so far on this project?

We're in the discovery and use-case identification stage for a pilot — there's no finished implementation to report yet.

### Why doesn't this case study have impact figures?

Because the project is active, not finished. We publish the process as it is, without filling in made-up results.

### How does this connect to the FOX project?

We used our AI training experience from FOX as a concrete starting point in the initial discussion — an already-delivered example, not a theoretical promise.

### Let's see what can be automated in your business.

A free 30-minute session: we'll tell you what can be automated, how long it takes and what it costs, with a fixed price after discovery.

We reply the same business day.