Project Name:


Project Description:

Mozaic combines AI technology and LayerZero’s omnichain technology to automate and optimize the yield farming process for the end user.

Core Team and Qualifications:

Calum Roberts: Founder (Core Contributor)

@MomentoMori1814: Head of Growth (Core Contributor)

@0xFortesque: Community Manager (Core Contributor)

John Rafman: Tech Lead/Senior Solidity Engineer (Core Contributor)

Tobi Jackson: AI Lead/Senior AI Engineer (Core Contributor)

Brendan O’Brien: Legal Advisor (Advisor)

Greg Tanaka: AI Engineer (Advisor)

@Kleronomas - DeFi and Economics Expert (Advisor)

@RAYBEAZY - BDM and Fund Management (Advisor)

Project Links:

Landing Page: https://mozaic.finance/

Docs: https://docs.mozaic.finance/

dApp: https://app.mozaic.finance

MOZ on Arbitrum: $0.04 | Mozaic Token (MOZ) Token Tracker | Arbiscan

Github: Mozaic · GitHub

X/Twitter: https://twitter.com/Mozaic_Fi

Contact Information:

Twitter: @0xfortesque

Email: 0xfortesque@proton.me

Telegram: @Oxfortesque


Requested Grant Size:

100,000 ARB

Grant Breakdown:

Solidity Development

  • 20,000 ARB

AI/ML Development

  • 15,000 ARB

Audit (w/ Trust-Security)

  • 35,000 ARB

Dynamically Adjusted, Time weighted Vault Incentives

  • 30,000 ARB

Required Development Additions:

  • Two Full-Time Senior Solidity Engineer

  • One Front End Developer

  • One Full-Time Senior AI Engineer

With the proposed ARB injections, the team will be capable of completing the Theseus vault by mid December.

Funding Address:


Funding Address Characteristics:

3/6 Multisig with all hardware wallets securely stored across different locations/individuals. We’ve started with 3/6 multi-sig initially and will upgrade to 5/7 signers (which is preferred) for more robust security measures in the near future.



The MozaicDAO is seeking 100,000 $ARB from the GMX Open Grants that will be used to advance its already in production Theseus Vault. With the proposed injection, the vault will be live to the public by mid-December bringing our already live AI ‘Archimedes’ to GMX V2.

OBJECTIVE #1 - Pre Audit: Finalize Development (commenced Early September)


  1. Development of Theseus Vault Contract incl. GMX ‘Plugin’

  2. Development of Control Center (CC) and Iron Hand (data hub)

  3. Testing with Control Center and AI

Current Progress:

  • Research conducted

  • Developed vault plugin for GMX

  • Developed Iron hand (data hub) for model training.

Estimated time to complete with additional resources: 10-14 days.

KPIs measured in Development Goals for OBJECTIVE #1:

  • Finalized vault contract and plugins for Vault

  • Fetching appropriate data from the pools (from DefiLlama) and integrating with the data hub (Iron Hand)

  • Codebase ready for Audit, pre-launch marketing campaigns begin

OBJECTIVE #2 - Post Audit: Beta and Public Launch


  1. Audit for developed smart contracts, syncing with marketing campaigns with partner protocols

  2. Deployment of smart contract on Mainnet w/ CC and Iron Hand

  3. Front end update and release

Estimated time to complete: 21 days


  • Initial audit report received for the Theseus Vault (duration ~15 days), marketing campaigns kick off with GMX

  • Mitigation round conducted and

  • Full Mainnet ‘Private Beta’ before ‘full public release’

OBJECTIVE #3 - Post Audit: Incentives

Periodic security vault cap increases

  • $10,000
  • $50,000
  • $100,000
  • $250,000
  • Up to 1M organic TVL before grant incentive distribution.

Dynamically adjusted based on performance and demand. We maintain a conservative approach to growth but are always able to accommodate according to demand. However, protocol security/vault security will always be of utmost importance to us and we plan to scale accordingly.

How will receiving a grant enable you to foster growth or innovation within the GMX ecosystem?

With its Theseus vault, the Mozaic protocol will bring its one of a kind AI technology to GMX V2 to attract additional liquidity providers to the platform and to the Arbitrum ecosystem. Theseus will bring one click, automatic yield farming FIRST to GMX V2 to align with its expansion plans on the back of its STIP approval. Every hour, the AI (Archimedes) will scan available yield across GMX V2 and if necessary rebalance (swap) assets so it can then allocate those assets to the highest yielding farm. If Archimedes does reallocate assets, it will also automatically compound.

We also use ‘Zap’ technology through Enso Finance which allows users to deposit into the Theseus Vault from almost any coin or LP token (including UNI V3 positions). This will further facilitate the acquisition of liquidity for the GMX V2 platform as we expand deposits from more chains.

Our flagship vault - Hercules - has seen rapid growth in TVL in its first two months acquiring ~$100,000 in organic, non-incentivised TVL. See DeFiLlama for more information: https://defillama.com/protocol/mozaic. Currently Hercules pays out approximately 7.5% - 10.5% in REAL YIELD. The AI calculates all farming decisions based on block by block data, and considers slippage, txn fees, liquidity, farm/pool reward share per block, price data and more to deliver the APY on the front end to users.

We’ve shown significant interest and TVL growth since inception and believe this TVL interest and growth will spill over to the Theseus Vault upon launch.

See ‘Protocol Roadmap’ section for information regarding Theseus TVL growth targets.

Justification for the size of the grant:

Mozaic is a community centric protocol and to date has taken NO VC INVESTMENT. The core contributors have built the entire AI technology and vault contracts from scratch. We request this grant to support development exclusively for the betterment of the Mozaic protocol and to further align with GMX and Arbitrum.

Grant Breakdown:

Solidity Development

20,000 ARB

  • Two Senior Solidity Developers (15,000 ARB)
  • One Front-End Developer (5,000 ARB)

AI/ML Development

15,000 ARB

  • One Senior AI engineer (12,000 ARB)
  • One Data Analyst (3000 ARB)

Audit (w/ Trust-Security)

  • 35,000 ARB

Dynamically Adjusted, Time Weighted Vault Incentives

  • 30,000 ARB

Grant Timeline:

FIRST payment of 35,000 ARB (35%%) to be released upon grant approval of grant to hire the additional engineers and bump the development timeline.

SECOND payment of 35,000 (35%) to be released upon achievement of OBJECTIVE #1 for payment of the Audit with Trust-Security.

THIRD payment of 30,000 (30%) to be released upon achievement of OBJECTIVE #2 and launch of the vault once 1M in organic TVL is achieved without any form of ARB token incentives.

We’ve structured the above grant release timeline akin to a development grant where the initial 35% of distributed ARB is exclusively focused on talent and product development for our Theseus Vault & are very confident that the release of the remaining grant tranches would reward GMX & ARB ecosystem users handsomely, supercharging our healthy organic TVL growth.

Fund Streaming:

The Mozaic protocol agrees to have funding of the proposed ARB grant streamed linearly for the duration of the grant proposal, and that the multisig holds the power to halt the proposed streamed grant at their discretion at any time.


What date did you build on GMX?:

Production for the Theseus Vault began in late August/early September.

Prior to this, we developed an automatic yield farming stablecoin vault ‘Hercules’ that has been live now for over two months, since August 23rd. Hercules is built atop Stargate Finance across seven chains including Arbitrum with ~$100,000 in TVL and growing.

To develop the Theseus Vault we borrowed lessons learned from building the AI and vault architecture for Hercules and have adopted a number of different approaches, which allows the team to streamline the vault development and efficiency of the AI itself. Here is previous testing of the Hercules vault: Machine Learning - Mozaic.

Protocol Performance:

The Mozaic protocol has taken NO VC investment to date. The AI technology has been in development for over two years with continuous training and improvements carried out concurrently.

Our acceptance into the Google Cloud Web3 start-up program ensures we have the resources to maintain, learn and enhance upon our current infrastructure (https://x.com/Mozaic_Fi/status/1699407136495981019?s=20)

On July 3rd, the protocol raised $300,000 from the community and DAOs alike - this sold out in less than 24 hours.(https://x.com/Mozaic_Fi/status/1676270607208697856?s=20)

In September, the protocol conducted an LBP on Fjord Foundry. Funds were used to continue development and bootstrap liquidity for the MOZ - ETH trading pair. The DAO owned liquidity, which now sits at over $700,000, is exclusively on the CamelotDEX V3 AMM with over $2,000,000 in swap volume. Over $60,000 in protocol revenue from LP fees have also been accrued in less than three months. There is also an upcoming partnership with Concentric.fi for our ALM provider, cementing the MOZ token within the Arbitrum ecosystem.

As a team and community, we have developed real working products that are already live. The grant would help the protocol expand and solidify our presence in the Arbitrum ecosystem. We see ongoing mutual synergies and growth with the launch of Theseus on GMX V2.

Protocol Roadmap:

Q4 2023

  • Theseus Audit report prior to launch
  • Theseus Vault Launch
  • Protocol Leaderboard (based on interactions with the protocol incl. Theseus)
  • Arbitrum Hoplite NFTs
  • Theseus security vault cap increase: $100,000 TVL target.

Q1 2024

  • Account abstraction and fiat on ramp for Theseus
  • Theseus security vault cap increases $500,000 - $1M TVL target, dynamically adjusted if/when demand increases.
  • Theseus security vault cap $1,000,000+ TVL target achieved and ongoing ARB incentives initiated.

Q2 2024

  • Theseus security vault cap increase to $5,000,000 by the end of Q2 maintained through superior offering of yield
  • Additional integrations of Arbitrum protocols within the Theseus vault - planned: Gains Network, Gamma Strategies and more.
  • Hercules ZK-AI, verifiably proving the off-chain model

Q3/Q4 2024

  • Theseus ZK-AI, verifiably proving the off-chain model
  • $10,000,000 + in ‘sticky’ TVL
  • Additional Arbitrum centric farms and protocols

Audit History:

Hercules Vault Audit: Audit - Mozaic

MOZ Token Audit: Audit - Mozaic

We are also partnered with Upshield to provide an additional, real time layer of alerts to the team to any suspicious/malicious activity: Upshield - Mozaic

SECTION 5: Data and Reporting

Is your team prepared to create Dune Spells and/or Dashboards for your incentive program?:

Yes, we are able to create our own Dune Dashboards for the incentive program and even the grant streaming itself through our partners at Hedgey.

Does your team agree to provide bi-weekly program updates on the GMX Forum thread?:


Additionally, we also host Community Calls every week with regular updates to our community and public.

Does your team acknowledge that failure to comply with any of the above requests can result in the halting of the program’s funding stream?:


Finally, on a personal note and regardless of the outcome of this proposal - we would like to extend our gratitude to the team from GMX, they have been extremely supportive as we have been building out the Theseus Vault.



Hey @fortesque, Thanks for applying for a GMX grant we will start reviewing the process and get back to you as soon as possible.

1 Like

I don’t know why my post was deleted!

To requote:

‘Thank you ser, feel free to ask any questions here in the meantime!’

1 Like

I hope that MOZAIC Finance gets the Grant


Mozaic is innovative and has a first mover advantage in the AI yield farming space. A grant can help them flourish which will be good for the GMX ecosystem.


Mozaic is the future of finance.


A very deserving project.


So optimistic about this project!

1 Like


AI-Optimized Yield and Liquidity Strategies | Powered by LayerZero :fire: :fire: :fire:

:heart: :heart: :heart:

1 Like

I have hope that MOZAIC Finance gets the grant!

1 Like

Mozaic is changing the world of DEFI. I strongly support their endorsement for this grant.

1 Like

Legit project, legit team

1 Like

Please support these guys! super hardworking team.

1 Like

Hey Team, unfortunately, we are rejecting your grant application. The proposal doesn’t align with GMX other than generating higher yield opportunities with V2 vaults if and when they arise. Also, a 100k grant request for a product with 100k TVL currently, also seems excessive. We would want to see a more significant impact on GMX, and more direct benefits, to feel it’s worthwhile.

Hi Saurabh!

First of all, I’d like to extend my gratitude towards the GMX team for reviewing the grants proposal & a decision, irrespective of the outcome.

I would like to add some important context here for a more nuanced discussion (as a community member foremost) :

  • Mozaic Finance was launched on 23rd August, 2023 on Arbitrum with the Hercules Vault (i.e. stablecoin yield vault) being first of many product offerings as a part of the project’s broader roadmap.

  • For broader comparision w peers, Hercules Vault was able to capture $100k USD in stablecoin deposits (which stands at 150k as of this date) within a matter of 3 months WITHOUT any native protocol incentives, as compared to equivalent product offerings for top 5 protocols in the same category (i.e. stablecoin yields).

For reference, Sommelier Finance, a top 5 protocol with a market-mover launch advantage by substantial amount still boasts only ~3M, which represents an extremely small portion of their total TVL capture.

Vault TVL Statistics (Non-exotic stablecoins)
DefiLama (Yield Aggregato/Optimisers)

  • The current TVL numbers are simply reflective of the protocols’ security-first approach that takes vault cap deposits & uninterrupted protocol functioning into active consideration., and it’s a “feature, not a bug” scenario to be precise.

  • Furthermore, the product team is certainly planning to launch Theseus Vault atop GMX v2 by mid December 23’/January 24’ nonetheless & purpose of the grant was simply intended to supercharge existing development efforts via focus on talent acquisition & security.

  • Additionally, the proposal seems to be structured in a way that the protocol DOES NOT seek any ARB incentives (30K ARB) UNTIL it has accomplished a miniumum organic TVL growth threshold of its own (~1M), as a way of showing continued committance towards planned growth objectives for a nascent protocol, as well as mission first being safety.

This, on its own, in my view makes it seem a much more attractive proposition than an excessive, immediate ask of the full $100K ARB amount.

The initial distribution percentages seems to be heavily front-loaded towards development efforts (65-70%, i.e. 70-75k), so I hope that a discusssion channel can still be kept open if there’s room for negotiation on :

  1. Bringing development costs down, AND/OR;

  2. Revised milestone targets,

  3. exploring potential mutual synergies.

Again, thanks for all your input & kindly do share any constructive feedback for the team (if any).


After some feedback from the GMX team and community, we would like to make an amendment to our grant ask and allocation.

Amended Total Ask: 75,000 ARB


1) Partial Payment for Audit with Trust-Security

• 10,000 ARB

2) Theseus Vault TVL Incentives PHASE ONE:

Released in-line with security vault cap increases after vault launch to kickstart growth

• 20,000 ARB

TVL CAP increases:

• $50,000 —> $100,000 —> $250,000 —> $500,000 —> $1M

3) Theseus Vault TVL Incentives PHASE TWO:

Released upon reaching $1M in TVL

• 45,000 ARB

Milestone based Farm incentives:

• 15,000 ARB distributed over 3 months

• 10,000 ARB when $2M TVL reached across Theseus GM farms

• 10,000 ARB when $3M TVL reached across Theseus GM farms

• 10,000 ARB when $4M TVL reached across Theseus GM farms


I hope this go trough after the changes.


it will be great for Mozaic


How does one apply for a grant

Hello Mozaic AI Team,

We want to thank you for applying to the GMX grants program. The window for applications has recently closed. In the interest of clarity, and to ensure continued lines of collaboration, we want to notify that your grant did not muster full support from the community to move on to approval. Such developments tied only to the STIP grant program are not the end of GMX working with you all. We understand that the opportunity presented by the grants program might have left you in the middle of some developments - we are more than happy to find other ways outside the grants program to see these developments through to a productive end, or to adjust scope accordingly in the interest of such ends.