We don’t predict the future,
we model the variance.
Research
How we research
Find out moreWhat we research
Three fields, one stance: we only believe what we can measure.
Markets
We don’t forecast prices. We measure how far they can swing, and we test signals with data instead of opinion. The focus is on Swiss and European markets.
- Do technical signals such as the RSI hold on new data?
- What does the order flow show before the price does?
Machine learning
We check where models for time series and patterns really do better than simple rules.
- When does a model beat a simple rule, and when not?
- How do you find patterns in spending without hurting privacy?
Applied AI
We measure how AI systems change the web and the work of small companies.
- Which AI crawlers read Swiss websites, and what do they take?
- Where do AI agents really save time in a small company?
How we research
- 1
Question
We start with a clear question that can be answered with yes or no at the end.
- 2
Data
We collect the data and write down where it comes from and what is missing.
- 3
Test
We test on data the model has never seen, in many scenarios. Also against our own idea.
- 4
Result
We show a range, not a single number. Even when the answer is no.
From research to product
For us, research is not an end in itself. What we learn goes straight into our tools and into every client project. So our products keep getting better.
Avtoris Research · Batch 1
Status of Batch 1: Opens 1 January 2027
Do real research. Earn your division.
Avtoris Research Batches is a selective, unpaid research program. You work in a small team on a real paper, get published with a DOI, and climb from Division 5 to Division 1.
How it works
Apply
A short form, about 20 minutes. We check the basics, not your CV.
Qualifier Sprint
48 hours, solo. One research question per track, public data, a report of up to 4 pages and code that runs.
Batch season
8 weeks in a team of 3 to 4, led by an experienced member. One 30 minute call per week, the rest is async.
Review Gate
Every paper is reviewed for method, reproducibility and clarity before it goes public.
Publish and rank
Accepted papers get a DOI. Your season points decide your division.
Divisions
| Division | Title | How you get there | What you can do |
|---|---|---|---|
| Division 5 | Title: Applicant | Pass the application check | Join the Qualifier Sprint |
| Division 4 | Title: Rookie | Pass the Qualifier Sprint | Contribute to a batch project |
| Division 3 | Title: Researcher | Contribute to a published paper | Choose projects, write peer reviews |
| Division 2 | Title: Senior Researcher | 2 or more papers, one as lead, top 30% of your track | Propose topics, judge the Qualifier, lead projects |
| Division 1 | Title: Principal Researcher | 3 or more papers, lead author with a strong review, top 10% of your track | Lead projects, sit on the Review Gate, get a reference letter |
You can move up one division per season. Divisions 1 and 2 are capped per track, so they stay rare.
Tracks
Quantitative Finance
Market microstructure, factors, volatility and backtests on public data.
Open in Batch 1Applied AI and ML
Evaluations of existing models, benchmarks and small, focused experiments.
Open in Batch 1Market and Data Research
Quantitative market research and data on the Swiss economy.
Later
What you get
- Your name on a published paper, with a DOI and your contribution roles.
- A season title, for example: Division 2 Researcher, Quantitative Finance, Batch 1.
- A verifiable certificate of contribution you can add to LinkedIn.
- Featured on avtoris.studio/research: the best paper of each season and its authors.
- A reference letter from the founders when you reach Division 1.
- Work with people at your level from many countries.
Rules
We never charge fees. Not for applying, not for joining, not for certificates.
- This is not a job, internship or course. There is no pay and no job guarantee.
- You must be 18 or older.
- You must be able to use GitHub, WhatsApp and Discord legally where you live.
- Papers are published under CC BY 4.0, code under MIT or Apache 2.0. You keep your credit.
- You never work on Avtoris client projects.
Batch 1 timeline
- Applications open
- 1 January 2027, 00:00 UTC
Further dates will be announced in December.
FAQ
Do I need a university degree?
No. We select on the work you do in the Qualifier Sprint. Students, apprentices, graduates and people on a gap year are welcome.
How much time does it take?
At least 6 hours per week for 10 weeks, plus one weekend for the Qualifier Sprint.
Can I use AI tools?
Yes, we strongly encourage it if properly documented. Reports without real work of your own are not scored.
What if I am not selected?
You get your scores. Strong submissions go on a waitlist for the next batch. You can always apply again.
What does it cost?
Nothing. We never charge fees.
Is the certificate accredited?
No. It is a certificate of contribution from Avtoris. The value is in the published paper behind it.
Apply for Batch 1
About 20 minutes. Please answer in English.
All fields are required unless marked optional.
Application received.
We will email you about the Qualifier Sprint.