How JellyC Uses Artificial Intelligence
Edited by Nick Mace
At JellyC, artificial intelligence (AI) is used to improve insight, efficiency, and communication — never to replace human judgement. AI is applied thoughtfully across research, execution, operations, and investor engagement, all within a controlled and responsible governance framework.
Our objective is simple:
Deliver better information, faster and more clearly, while upholding the standards expected of a regulated investment manager.
1. Firm‑Wide AI Governance & Control Framework
JellyC operates AI under a formal, board‑approved governance structure — not ad‑hoc experimentation.
JellyC’s governance stack includes:
Board‑level AI Use Policy
Firm‑wide AI Standard & Procedure
AI Trading Control Standard (specific to investment and execution use)
These documents define:
Permitted vs. prohibited AI use
Risk classifications and approval thresholds
Mandatory human oversight
Confidentiality and information‑handling rules
Incident escalation, monitoring, and record‑keeping requirements
For any material AI‑supported output, the firm must be able to explain:
What system was used
The specific role AI played
Data class involved
Degree of autonomy
Human review performed
Known model or system limitations
Implication:
AI at JellyC is treated as a regulated capability, not a generic productivity tool.
2. AI in Investment Research & Portfolio Intelligence
JellyC uses AI and machine learning to enhance research, monitoring, and decision support — not to run fully autonomous portfolios.
Internal research initiatives include:
AI/ML models for portfolio and market monitoring
Real‑time data ingestion pipelines
Analytics and predictive models assessing market trends and asset behaviour
AI is used to:
Process large volumes of market and on‑chain data
Generate quantitative insights that support portfolio decisions
Assist in drafting research notes and investor communications
In practice, JellyC monitors multiple indicators (e.g., sentiment, reserves, technical indicators) and uses AI models for tactical insights — with human managers retaining decision authority.
Implication:
AI augments, rather than replaces, portfolio managers.
Turning Research Into Audio: AI Podcast Hosts Bruce & Tilly
To improve accessibility, JellyC uses AI podcast hosts — Bruce & Tilly — to convert research articles into professionally narrated audio.
This allows investors to:
Listen to research on the go
Understand complex topics through conversational explanation
Stay informed even when time‑constrained
Bruce & Tilly do not generate independent views; they simply present existing JellyC research in audio format.
3. AI‑Supported Trading & Execution (Controlled Use)
Where AI interacts with trading or execution, JellyC enforces additional layers of control.
All AI‑managed or AI‑initiated trading must comply with the AI Trading Control Standard, which governs:
Approved, validated data sources
Model governance and prompt controls
Access security and credential protection
Digital‑asset‑specific market integrity protections
Live monitoring, incident management, and kill‑switch protocols
Leadership is trialling agentic execution, but only within a proprietary, tightly controlled execution infrastructure.
Implication:
AI‑driven trading is allowed — but only inside a rigorously defined and monitored perimeter.
4. AI in Operations, Reporting & Internal Productivity
AI also supports operational efficiency, with strict attention to privacy and confidentiality.
Documented use cases include:
AI‑based note‑taking and summarisation (with deletion and privacy controls)
Internal guidance on acceptable AI usage
Ongoing internal initiatives:
AI‑enabled acceleration of NAV and reporting workflows
Enhanced analytics and reporting tools
Internal AI use is governed by policy, not left to individual discretion.
5. AI in Client & Public Engagement
JellyC uses AI externally in controlled and transparently disclosed ways.
AI Assistant “Bluey”
A website‑based assistant that helps users:
Answer general questions about crypto markets and JellyC funds
Improve information accessibility
Public disclosures clearly state:
Responses are AI‑generated
Outputs may contain errors
Content is informational and not investment advice
Implication:
Client‑facing AI is informational only, not advisory.
AI‑Powered Due Diligence (DD) Bot
JellyC also supports investors via an AI‑enabled DD bot, designed to:
Help users navigate JellyC materials
Answer standard & complex DD questions using approved datasets
Surface relevant documents quickly
Final clarifications and confirmations remain available through the human JellyC team.
6. AI as a Strategic Capability
Senior leadership positions AI as part of JellyC’s competitive edge, alongside proprietary technology and institutional market experience.
JellyC’s strategic identity combines:
Proprietary technology
AI‑powered research and analytics
Institutional‑grade fund structures
This mix enables JellyC to exploit structural inefficiencies in digital‑asset markets.
7. Agentic AI Use at JellyC
JellyC recognises that certain AI systems are capable of executing sequenced actions, using tools, interacting with external information sources, and operating with a defined degree of autonomy. JellyC may deploy agentic AI only where such use is consistent with the firm’s governance framework, risk appetite, and control environment.
Because agentic AI can initiate and progress actions rather than merely assist, JellyC treats it as a higher‑governance category of AI use.
In support of this governance approach, JellyC is preparing to deploy structured agentic workflows, using Crystal aOS platform, to enhance productivity while embedding compliance pathways directly into tasks performed across the organisation. These workflows are designed to sequence actions, approvals and control checkpoints so that policy, risk, and compliance requirements are addressed as part of task execution, with human accountability and decision ownership retained at all times.
Any agentic AI deployment must operate within:
Clearly defined authority and scope
Explicit technical and operational boundaries
Documented human accountability and supervision
Agentic systems are subject to:
Formal approval prior to deployment
Ongoing monitoring and traceability
Escalation and override mechanisms
Controls proportionate to task sensitivity and error impact
Agentic AI may not operate as an unconstrained autonomous actor on behalf of JellyC. It may function only within specifically approved mandates and controlled environments.
Where agentic AI interacts with:
Trading or execution systems
Wallets or treasury functions
Exchange APIs
Other sensitive operational or financial systems
It must comply with the AI Trading Control Standard and all applicable security, operational, and compliance requirements.
Implication:
Agentic AI at JellyC is treated as a strategic but tightly governed capability, requiring enhanced controls, explicit accountability, and stricter oversight than conventional AI‑assisted tools.
Summary
JellyC uses artificial intelligence as a governed, risk‑controlled capability across research, portfolio intelligence, execution, trading support, operations, and client engagement. AI enhances data analysis, monitoring, and decision support, while all material outputs remain subject to human oversight. Trading‑related AI operates under a dedicated control standard, and all AI use is governed by board‑approved policies covering confidentiality, explainability, risk classification, and incident management. Client‑facing AI tools are informational only, with clear disclosures stating they do not provide investment advice.
Disclaimer
This article ("Article") has been prepared for informational purposes only and does not constitute an offer to sell or a solicitation of an offer to purchase any financial product or service. This Article does not form part of any offer document issued by JellyC Pty Ltd (CAR Number 001293184), a corporate authorised representative of TAF Capital Pty Ltd (ACN 159 557 598, AFSL 425925). Past performance is not necessarily indicative of future results, and no person guarantees the performance of any financial product or service mentioned in this Article, nor the amount or timing of any return from it.
This material has been prepared for wholesale clients, as defined under Sections 761G and 761GA of the Corporations Act 2001 (Cth), and must not be construed as financial advice. Neither this Article nor any offer document issued by JellyC Pty Ltd or TAF Capital Pty Ltd takes into account your investment objectives, financial situation, or specific needs.
The information contained in this Article may not be reproduced, distributed, or disclosed, in whole or in part, without prior written consent from JellyC Pty Ltd. This Article has been prepared by JellyC Pty Ltd, which, along with its related parties, employees, and directors, makes no representation or warranty as to the accuracy or reliability of the information provided and accepts no liability for any reliance placed on it. Prospective investors should obtain and review the relevant offer documents before making any investment decision.
