This draft records a regional and title-blind discovery pass across conference programs, firm pages, publisher episodes, academic media, and local-language surfaces. It widens the search beyond English-language shows whose titles explicitly mention hedge funds or AI. The records below are public-signal evidence; they do not rank firms or establish a production deployment unless the source says so directly.
Episode URL: Multiple canonical episode, conference, firm, academic, and platform URLs are linked in the tables below; this is a multi-source ledger, not a single-episode transcript.
Source files: Public publisher, firm, conference, academic, and platform pages linked inline below; the internal capture ledger is maintained separately.
Source status: Regional discovery and verification pass. Audio recovery, transcript checking, identity cross-checks, and replay access are incomplete for some records.
Credibility: HIGH for canonical page identity and stated metadata; MEDIUM for first-party descriptions and guest accounts; LOW for any claim not supported by the linked primary surface.
Evidence Boundaries: Conference metadata, publisher biographies, firm claims, and academic media are kept separate from evidence of a named manager’s live system, permissions, model performance, or returns.
Verified additions
| Surface | Public source | What is observable | Boundary |
|---|---|---|---|
| Infinity Global Asset Management / Jongwon Roh | Korea Investment Week 2026 program · replay/material locator | The Korean/English program lists Roh as CIO and gives a session titled “Pioneering the Next Generation of Korean Hedge Fund Management through AI and Quantitative Strategies.” | Official program metadata. Replay/material access was gated in this pass; no transcript, slides, model inventory, or performance result was available. |
| Asset Management One / AMOAI | Japanese Hedge Fund Eye Vol.45 PDF | The May 2026 note by Kohei Hayashi explains statistical-arbitrage process design and maps NLP to text features, ML to signal creation/combination, reinforcement learning to execution and portfolio construction, and deep learning to high-dimensional inputs such as satellite imagery. | First-party strategy note; no model weights, training corpus, permissions, or independently attributed returns. |
| OTHOZ Capital / Julien Florian Jensen | Hedgework Talk publisher page · Podigee episode · OTHOZ | The German publisher identifies Jensen as Executive Director and describes OTHOZ as a Berlin-founded firm using AI technology in investment products since 2018. | Publisher and firm description; no model names, training data, permissions, or independently verified results. |
| Etna Research / Marco Jean Aboav | Investology episode · Etna first-party site · Aboav’s public post | The episode identifies Aboav as an ex-hedge-fund manager and Etna CEO and covers data scouting, data engineering, quant pods, alpha discovery, and adaptive strategies. Etna names Aboav, David Ko, and Davide Panelli and describes frontier AI for institutional clients and asset owners. | Public interview, company, and social-post evidence. Product, speed, and cost claims remain company claims; no client identity, benchmark, data right, or return attribution. |
| High-Flyer Quant / DeepSeek / Liang Wenfeng | 36Kr Chinese interview · English translation and context · The Paper historical account | The 2023 Chinese interview identifies High-Flyer as an investor in a separately established DeepSeek company and says the founder viewed the AGI work as distinct from finance. It describes an AI-heavy team and sustained investment in model, data, and compute research. The Paper adds historical claims about an early AI trading model, GPU-based execution, and large training clusters. | Primary interview plus secondary historical reporting. The sources do not disclose High-Flyer’s current financial model inventory, whether DeepSeek research is used in trading, current personnel allocation, data rights, or audited investment outcomes. |
| High-Flyer Quant — current official quant surface | High-Flyer Quant | The firm’s English page says it began exploring fully automated quantitative trading with machine learning in 2008; it describes a first deep-learning trade in 2016, full deep-learning implementation in 2017, its Fire-Flyer AI deep-learning platform, more than 10PB of data, and thousands of continuously updated sources. It also describes neural-network and NLP use. | Firm-reported current positioning. The page does not provide model cards, data permissions, current strategy inventory, personnel ownership, or independent performance evidence. |
| JoinQuant Investment | JoinQuant AI-research page | The Chinese first-party page identifies JoinQuant Investment as an AMAC-registered private-fund manager founded in 2017. It describes AI-driven investment research across factor, model, optimization, and decision layers, and says the team includes ICML/ICLR publishing researchers. | First-party positioning; no named live model, paper-to-production mapping, vehicle list, data-rights disclosure, or independent evaluation was recovered. |
| Ubiquant Investment | Clocktower China Quant Deep Dive | The 2024 report says Ubiquant established Data, AI, and Waterdrop labs for data, algorithms, and trade execution. | Secondary report only. Current legal status, personnel, lab activity, model inventory, and production authority remain unresolved. |
| SpringPad / SpringPad Wealth Solutions | Hindi-facing workshop · login-gated course page · public intro media page · public “20th unedited” media page | A title-blind search found a separate first-party learning site that exposes a course titled “1 Day Stock Market Using AI – Workshop: Session Recording and Notes,” but the course requires login. Two public media pages expose filenames without a transcript or verified downloadable content. | Metadata-only recovery lead. The landing page says recordings are not provided; the accessible media pages do not prove that the files are the July 19 workshop, are downloadable, or contain Hindi content. No substantive workshop claim is promoted. |
| China fund-industry AI research/risk forum | AMAC notice | The Asset Management Association of China announced a November 26, 2025 Shenzhen exchange on AI in fund investment research and risk control, with agenda items covering human-machine collaboration, public-fund research management, risk-control applications, and an AI-enabled compliance-risk platform. | Industry-level conference evidence, not evidence that any named manager implemented a system. The notice does not publish attendees or recordings; recover those before attributing statements. |
| France / Ploovers / Franck Béon | Les Investisseurs 4.0 episode | The French-language page describes Béon’s quantitative-hedge-fund background, multi-billion-euro ETF portfolio experience, and Ploovers’ use of mathematical models, AI, and empirically augmented data. | Guest biography and founder description; no prior-fund name, model architecture, training data, or live authority. |
| Korea University Financial Technology Lab | Official lecture index · LLM investment analysis · AI versus financial analysts · ML in asset management | The lab’s Korean/English index exposes media on LLM investment analysis, AI versus analysts, machine learning in asset management, portfolio optimization, scenario analysis, and backtesting. | Academic media, not evidence of a named fund’s deployment, client data, or investment result. |
| Kinea Investimentos / Kinea Insights | Kinea first-party agent essay · Kinea Insights episode · Kinea team page | Portuguese first-party media discusses autonomous agents, tool use, productivity, business-model exposure, and implications for Kinea’s multimarket funds. The team page names Mariana Marques Smidt, Rodrigo Zobaran, and Ruy Alves in data-science, quantitative-research, AI-infrastructure, and multimarket roles. | First-party commentary and personnel evidence. No complete agent inventory, model permissions, training data, production authority, or independently attributed outcome is disclosed. | | Kinea / Ruy Alves | Os Economistas episode · Market Makers episode · XP Expert Talks | Portuguese-language media identifies Alves as a Kinea multimarket manager and discusses AI-market structure, DeepSeek, automation, valuation, and investment positioning. | Public interview and publisher evidence; no internal model inventory, data rights, live decision authority, or performance attribution. | | Itaú Asset Quantamental | Itaú Views episode · Itaú quantitative-competition release | The Portuguese episode names Victor Dweck and Pedro Barbosa in a discussion of quantitative funds, investment robots, AI, and human capital. The first-party release documents a quantitative-robot competition and a finalist project using machine learning. | Asset-manager and talent-pipeline evidence; no specific live model, governance, permission, or independently measured result. | | Noax Global / Ivan Blanco | Zona Quant episode · recovered Ivoox enclosure · capture ledger · Spanish fund article | Recovered audio and local Spanish ASR identify Blanco’s quantitative background, factor-investing framework, Noax’s systematic equity/volatility/fixed-income components, sector diversification, and a human fundamental-review step after systematic selection. A separate Spanish finance article describes Noax as combining factor investing with machine learning and AI. | Dated practitioner account plus secondary publisher reporting. Current status, model architecture, data, permissions, and performance remain unresolved. | | Neo Ivy Capital / Renee Yao | Odds on Open episode · SEC adviser record · Neo Ivy profile · public AI article | The episode presents Yao’s former Citadel/Millennium roles and Neo Ivy’s self-described AI/statistical-arbitrage process. SEC and LinkedIn surfaces resolve the adviser/entity identity and public deep-learning positioning; the company-hosted article adds named authors. | Promotional AI and AUM descriptions are not independently validated; no current architecture, permission map, or performance attribution is disclosed. | | Equity Data Science / Sandeep Varma and Benjamin Lieblich | Momentum episode · EDS leadership · EDS events | The episode and first-party pages connect Varma’s former buy-side quantitative roles with a platform that joins research, portfolio construction, risk, attribution, and AI. The event archive exposes additional AI-driven risk-management and quantitative-workflow discussions. | Vendor and founder evidence; no tracked-fund deployment, customer-specific data rights, model evaluation, or investment outcome is established. | | Wright Research / Sonam Srivastava | The India Opportunity · Wright first-party site · Wright team/product surface · AI/ML fund material | Indian quant/PMS media and first-party surfaces identify Srivastava as founder and portfolio manager, with HSBC, Edelweiss, and Qplum experience. They describe factor and regime modeling, AI/ML forecasting and allocation, automation with human oversight, CTO Vinod Reddy Kotha, and ML adviser Dr Miquel Alonso. | First-party biography and product claims plus podcast metadata; no model architecture, training data, permissions, live authority, or independently audited result is established. Marketing figures remain unverified. | | JioBlackRock / Rishi Kohli | The Brand Called You episode · IIQF podcast index | India-facing media identifies Kohli as JioBlackRock CIO and discusses quant strategies, alternative data, systematic models, and Aladdin; IIQF exposes a broader AI/ML-in-finance podcast surface. | Public interview and institute index; no JioBlackRock model inventory, data rights, production permission map, or performance attribution. | | Saudi Exchange Algorithm-Enhanced Trading Fund / Saudi Fransi Capital and Winton | Arabic Saudi Exchange prospectus · capture note | The Arabic prospectus documents an algorithm program for return prediction, risk prediction, portfolio construction, and cost control. It names Saudi Fransi Capital as fund manager and Winton Capital Management Limited as investment adviser, with responsibilities for algorithm development/operation, data processing, target portfolio weights, and program updates. | Strong regulatory evidence of a stated operating design; it does not prove live scale, performance, model weights, training data, or current implementation beyond the prospectus terms. | | JCube Capital Partners | JCube first-party firm page | Singapore-facing first-party material describes a MAS-licensed systematic fund manager using statistical/empirical models, machine-learning signals, alternative data, and large-language-model outputs in its research framing. | First-party positioning and licensing statement; no fund names, model architecture, training data, permissions, or independent outcomes. Verify the licence through MAS. | | Aggregate Asset Management | Machine learning and investing · 2026 announcement | Singapore first-party pages describe machine learning for stock selection and a later expansion of the proprietary “Deep Deep” model in a flagship fund. | Firm-reported process and model naming; no weights, evaluation protocol, permissions, or independent attribution. | | Avangard Investments / A.L.F.R.E.D. | Australian strategy page | The page describes an adaptive-learning system for ranking equities and derivatives, a systematic Australian-equity fund beginning July 1, 2026, traceable data/rule outputs, and investment-team review. | First-party system description; no implementation details, independent results, data rights, or GenAI-specific component. | | STANLIB Systematic Solutions / Chetan Ramlall | Moneyweb interview · Citywire audio mirror | South African publisher media identifies Ramlall as Head of Quantitative Research and discusses machine learning, alternative data, satellite/drone imagery, automation, and risk quantification. | Publisher interview and guest account; no current model inventory, permissions, or independently measured result. | | NMRQL Research / Stuart Reid | TWIML episode | The dated South African episode identifies Reid as Chief Scientist and describes adaptive, testable trading decisions using ML/DL for non-stationary financial time series. | Historical publisher and guest evidence; current status, personnel, model versions, and live results require re-checking. |
Reviewed but not promoted
Chinatown 2.0’s Robbie Yan episode is now cross-checked against Metacast, Podcast Republic, a Chinese LinkedIn profile, and a The Org profile. The publisher surfaces agree on the 2020 episode identity and guest description; the LinkedIn and The Org pages do not safely establish that they are the same person or identify the fund discussed in the episode. This remains a China-focused personnel and metadata lead, not current firm evidence.
Evidence status
The full internal capture ledger and coverage metadata remain separate from this public note. The latest Singapore/Australia/South Africa capture details are recorded in the regional manager sweep note. Audio recovery, transcript verification, identity cross-checks, and replay access are incomplete for some records. Every claim above is limited to the linked public surface and its stated boundary.