Blog · Published · Updated
Local AI search for company presentation slides
Short answer
A local slide search finds a slide from a description of what the employee remembers and shows the thumbnail, the presentation's name and the pages around it. The employee opens the original and decides whether the material is current and may be shown to a client. At Syntalith an app on a local model costs from €3,000 net plus GPU hardware. At 400 slide searches a month of 10 minutes each, it pays back in 2.0 months.
How we workDetails ↓
- 01Call30 minutes with an engineer, free
- 02PrototypeA working system on your process and data, free
- 03ImplementationTarget and fixed price in the contract; 50% at signing
- 04Acceptance50% once the agreed target is met
- model on your server or in the cloud in the EU
- quoted after the review
- same work in application with EU cloud model
- from €3,000
- task, data and equipment analysis
- €0
We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓
Author: Syntalith · 3 min read
Statistics
Searching company documents: statistics 2026
2.0 mo.
payback of a €3,000 local app at 400 slide searches a month: €2,327 of work today, €796 with checking and a GPU server. Scenario.
49%
According to Anthropic (2024), combining contextual embeddings with BM25 search cut the number of failed retrievals by 49%.
48%
According to the European Commission (2026), 48% of EU SMEs and mid-caps name a lack of expertise as a barrier to AI.
Source: European Commission, Flash Eurobarometer 578 (2026) ↗
20.0%
According to Eurostat (2025), 20.0% of EU companies with at least 10 employees used AI.
15.1%
According to Eurostat (2025), 15.1% of people aged 16 to 74 in the EU used generative AI tools for work.
Figures checked: 22 October 2026
A chart you found needs its context
A consultant is looking for a presentation about ways of organising customer service. They remember a chart comparing one shared team with service split between branches, but they do not remember the file name. They search folders and open one presentation after another. The search finds a slide that matches the description and shows it with the presentation's name, the date and the pages around it.
An image you found does not confirm that the data is current or that you may show it to another client. The owner of the archive sets which material may be reused. A presentation may hold the chart as an image and the explanation in a caption or on the next page, so the search reads those parts too.
- A thumbnail of the whole slide
- The presentation's name, date and author
- The pages around it with the explanation
- A link that opens the original
- Whether the material may be shown to clients
What local slide search costs and when it pays back
- Today
- 400 slide searches × 10 min = 66.7 h × €34.90 = €2,327
- With the app, an employee checks the output in 25% of the time
- €582
- Hetzner GEX45 GPU server
- €214
- Saving
- €1,531 a month
Payback of €3,000
2.0 months
The volume, the time and the checking share are the scenario's assumptions, which we measure in the trial. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat. Price of the GEX45 with 24 GB of GPU memory from Hetzner's website on 22 October 2026, without the one-off setup fee of €209.
The count covers the time spent searching, reading and compiling the material. What stays is the time to check the result against the source, 25% of today's work in the scenario. We measure both times in a trial on 20 to 100 cases the team has already handled.
Copies made for search are data too
The search makes copies of the text and thumbnails of the slides. If the material has to stay in your own environment, that also covers these copies, the index and the search history. We agree the sources and the rules for reuse with the archive's owner, and the environment and the handling of changes to the collection with IT.
- Where the app takes documents from and where it saves results
- What stays in logs, backups and the history of questions
- Who has access to questions, answers and summaries
- Who updates the model, the app and the set of documents
Tidying the archive may be enough
If the problem is files on private drives, you first need a shared place and rules for filing material. A few well-described template presentations can help more than searching the whole history. Docling describes support for PPTX presentations when processing documents. A search across internal material and a catalogue of slides approved for use with clients may cover different files, and the model does not make that choice for you.
Files on private drives
a shared place and filing rules
no app
A few template presentations
a catalogue of described material
tidying
Thousands of slides in the archive
local slide search
from €3,000
Where to start
The trial covers one collection of presentations and questions from employees who know what is in it. They compare the results with ordinary search and judge whether the preview lets them quickly reject a slide that looks similar and is of no use. We roll out local models as part of private LLMs, and we build the view for working with sources as an AI app. For the free process review, a description of one case and the places where the material sits is enough. Confidential files stay in the company, and you get a written plan within 2 working days.
Worked example · Scenario
How we count the payback: a presentation archive
We run the same count for your archive in the free process review.
€2,618
a month for the work on 300 questions
EU cloud version payback: 1.4 mo.
Our assumptions are 15 minutes to find an earlier agreement in the documentation, and checking in 15% of the old time. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat, for companies with 10 or more employees. Model cost: Claude Sonnet 5.5, about 6,000 + 500 tokens per answer in the EU cloud version, at the ECB rate of 1.1205 USD per euro on 22 October 2026. We count the payback for an app with a model in the EU cloud (from €3,000 net). We price the model on your server after a call, and the cost of people's work stays the same.
See the full calculation →
