Post-event report
AI in Finance - Paris
Where Finance Meets Artificial Intelligence
115 people from 91 firms held a place; 5 firms paid for the evening, so no one in the room had to.
Organised by
Who was in the room
115 people held a place
91 distinct firms
23 speakers on stage, from 22 firms
Roster, not a door count. Paris predates first-party registration, so nobody was scanned on arrival: 115 people held a place, and we do not claim 115 walked in.
115 registered · 1 excluded · 114 described. The record taken out is 1 duplicate registration. Every figure below describes the remaining 114. Attendees are only ever removed as duplicates, as our own or a host's staff, as records that are not people, or as records carrying no employer at all — never for being the wrong kind of attendee, and never for which firm they work for.
Who they worked for
×3 
×2 
×2 
- BNP Paribas CIB
×2 
×3 - Institut Louis Bachelier ×2
×4 
×3 - Cedrus & Partners ×2
×2 
×2 
×2 
×2 - Omaha Insights ×2
×2
What kind of firms
Held a place 115
Excluded from this report −1
Described here 114
Named no employer −3
Named an employer 111
No firm type held −3
Firm type on record 108
The room
n = 108 classified attendees
- Buy-side 24 22%
- Sell-side 9 8%
- Technology & data 44 41%
- Other sectors 31 29%
The stage
n = 22 speaker firms
- Buy-side 5 23%
- Sell-side 3 14%
- Technology & data 13 59%
- Other sectors 1 5%
22% buy-side — 24 of the 108 classified attendees: Asset management — 19 people from 17 firms; PE — 2 people from 2 firms; Hedge fund — 1 person from 1 firm; Sovereign wealth fund — 1 person from 1 firm; VC — 1 person from 1 firm.
The programme
15sessions
23confirmed speakers
22speaker firms
3h 55mof sessions
- Talk 1h 20m 22%
- Demo 1h 15m 21%
- Networking 1h 10m 19%
- Panel 1h 10m 19%
- Break, Ceremony 1h 10m 19%
| 25m | Arrival & Registration break | |
| 5m | Opening Remarks ceremony | 1 spk |
| 20m | Industry Talk: Decomposing, Forecasting & Trading Inflation — Using Machine Learning & Alt Data talk | 1 spk |
| 15m | Demo 1: The end of the wealth management advisor — or the beginning of their revolution? demo | 1 spk |
| 10m | Demo 2: Leveraging AI to drive Advisor efficiency and client engagement demo | 1 spk |
| 20m | Industry Talk: Machine Learning and AI at EvarInvest AM talk | 1 spk |
| 10m | Demo 3: Databricks & Neuralk demo | 2 spk |
| 20m | Industry Talk: Reading Microstructure with Temporal Fusion Transformers talk | 1 spk |
| 20m | Networking Break break | |
| 20m | CBOR and Semantics — a recipe for AI-ready client insights talk | 2 spk |
| 10m | Demo 4: Aisot Technologies demo | 1 spk |
| 10m | Demo 5: Zanista AI demo | 1 spk |
| 40m | Panel 1: Wealth Management x AI panel | 6 spk |
| 15m | Networking Break break | |
| 10m | Demo 6: Linkup demo | 1 spk |
| 10m | Demo 7 — Virgo: A Purpose-Built Language Model for Autonomous AI Agents demo | 1 spk |
| 30m | Panel 2: AI Agents in Finance panel | 5 spk |
| 5m | Closing Remarks ceremony | |
| 70m | Networking & Drinks networking |
Speakers
- Arman Khaledian Co-Founder & CEO Zanista AI
- Boris Toledano COO & Co-Founder Linkup
- Ludovic Bodin Founder & CEO BOBIC Generational Wealth
- Michael Althof CEO/CIO LEIW | Luxembourg Enterprise for Investment and Wealth Management
- Yannick Raynard Group Chief Technology Officer Harvest Groupe
- Friedhelm A. Schmitt CEO Fincite
- Fayssal El Mofatiche Founder & CEO Finteda
- Dr. Till Blesik Director DACH / Enterprise Architect Aiviq
- Saeed Amen Co-Founder & CTO Turnleaf Analytics
- Hanane Dupouy AllianceBernstein
- Nicolas Obolensky Chief Product Officer Harvest Groupe
- Iordanis Kerenidis Co-founder & CTO Quantum Signals
- Alexandre Cesari Director EvarInvest AM
- Caio Natividade Head of Quantitative Investment Solutions Research Deutsche Bank
- Nico Baum Head of Solutions | Head of AI Berenberg
- Laurent Fabre Field CTO, South EMEA Databricks
- Emmanuel Hauptmann CIO, Head of Equities. RAM Active Investments
- Stefan Klauser Co-Founder & CEO Aisot Technologies
- Jacob Chanyeol Choi Co-Founder & Co-CEO LinqAlpha
- Aaron Stillwell Field Engineering Lead Neuralk
- Surender Uttamsingh SVP, Senior Sales Director, Wealth Management - EMEA & APAC FactSet
- Gregory Viet Head of Client Operations HSBC
- Bertrand K. Hassani, PhD CEO Quant AI Lab
The evening
18 photographs published for this event — the first is the one at the top of this page. These are the same files the public gallery serves, so every frame can be checked against finteda.com/gallery.
Who backed it
23 organisations put their name to this evening. 5 of them paid for it; that is why the 115 people above paid nothing.
Sponsors 5
Partners 4
Organised by 1 Finteda
Contributors 13 FinciteFactSetHSBCTurnleaf AnalyticsAiviqQuantum SignalsDeutsche BankBerenbergRAM Active InvestmentsLEIW | Luxembourg Enterprise for Investment and Wealth ManagementEvarInvest AMLinqAlphaQuant AI Lab
Sponsoring the next edition works the same way — events@finteda.com.
Every figure here is a row count
Taken from Finteda's own systems when this page was generated. Nothing on it is a target, a projection, or carried over from another event.
- Excluded — 1 of the 115 are not described here: 1 duplicate registration. The counts above are over the remaining 114. No attendee can be removed for who they are, or for which firm they work for.
- Buy-side — capital being deployed, not intermediated: Hedge fund, Asset management, Pension fund, Family office, Sovereign wealth fund, PE, VC.
How these numbers were counted
- Roster — attendee records held for this event, deduplicated per person.
- Excluded — 1 of the 115 roster records is not described here: 1 duplicate registration. The counts above are over the remaining 114. An attendee may be removed only as a duplicate of another record, as Finteda's own or a host's staff, as a record that is not a person — a shared mailbox or an import artefact — or as a record for which no employer was ever captured. There is no reason available for removing a genuine attendee because of who they are; the employer reason names an EMPTY field and never a firm, so no attendee can be removed over which employer they have. Every removal is recorded against the person who made it. Where an exclusion applies it applies to THIS event only.
- Distinct firms — unique employers on that roster, case- and punctuation-insensitive, after three corrections applied by rule rather than by hand. A trailing legal form or generic descriptor (Ltd, SAS, GmbH, "Technologies", "Group", "Holdings", "International") is folded onto the parent name, so one firm written two ways is counted once. Words that genuinely distinguish two entities — Capital, Partners, Management, Global — are never folded. A trailing qualifier describing the person rather than the employer ("— Student", "— Intern") is dropped. And an answer that states an employment status or a stage instead of naming an organisation is not counted as a firm at all; those people stay in every other number on this page, because they attended.
- The firms named above — every firm on the roster whose company record reports 1,000 or more employees, or that sent 2 or more people to this event. 32 of 91 qualify. Each is named as its company record spells it, which is also where its logo, headcount and website come from; where we hold no logo the tile is set in type, and where we hold no website it is not a link. Firms below that line are not published, and no attendee is ever named, titled or counted individually on a shared copy of this page.
- Firm type — the employer's company record, matched by CRM link or exact name. 108 of 114 attendees resolve to a record carrying one; of the 6 that do not, 3 named an employer we hold no type for and 3 named none at all. That second group is the ceiling rather than the shortfall: a roster row naming no employer can never be classified by anyone, so 111 is the most that could ever have been reached here, and 108 of those 111 were.
- Buy-side — capital being deployed, not intermediated: Hedge fund, Asset management, Pension fund, Family office, Sovereign wealth fund, PE, VC. No other type counts, whether or not it appears in this room; banks, investment banks, brokers, exchanges, market makers and vendors are counted out.
- The four sides — a regrouping of those same firm types, not a second classification of the same firm. Every type lands in exactly one side and the four are exhaustive over the classified base, so the shares add back to it. Sell-side is Bank, Investment bank, Private bank, Broker, Exchange, Market maker, Trading. Technology & data is Fintech, Tech, Data provider, AI / ML. Other sectors is every remaining type — consulting, insurance, academia, regulators and the public sector among them: real parts of this room, and none of them a buyer, an intermediary or a vendor. It is named for the SIDE and not for the firm type also spelled "other", which is one of the types inside it.
- The stage — speaker firms are counted the same way attendee firms are, so two speakers from one firm are one firm. 22 firms put someone on this stage and 22 of them carry a type; the stage figure is drawn over those 22, in firms, against a room figure drawn in people. The two share a definition, not a base.
- Not reported — seniority and home city. 2 of 114 attendee records carry a seniority value and 43 carry a city, which is too thin a base to describe either, so this report describes neither and names no value for them. 112 do carry a job title, but bucketing titles into seniority would be a classification we invented rather than one we hold.
- Speakers — confirmed only; proposed and declined names are excluded.
- Backers — each organisation appears under the role the CRM recorded for this event, and sponsors are grouped by the package recorded against them. The tier name and its entitlements are quoted from that record; none of the sponsors on this event has one, so the section names who backed the evening and states that what they bought is not recorded — it is never inferred from a logo or from the order the marks appear in. No amount appears on this report. Marks and links are the company record's own, resolved exactly as the firm wall above resolves them.
- Sessions — agenda items excluding registration, breaks and networking. Durations are as SCHEDULED and were never observed, so no figure here claims when anything actually started; the allocation bar splits those scheduled minutes by item type, and its segments sum to the door-to-door total.
- Photographs — the published gallery for this event, served from finteda.com.
Every report we send is built from the same query as this one. If a number here does not survive your checking, tell us — that is the only kind of correction this document takes.
Generated 2026-08-07 07:14 UTC.