Anti-Scam Job Board w/ CIP-68 and AI-Led Screening - Retrai

[Proposal setup] Proposal title

Please provide your proposal title

Anti-Scam Job Board w/ CIP-68 and AI-Led Screening - Retrai

[Proposal Summary] Budget Information

Enter the amount of funding you are requesting in ADA

45000

[Proposal Summary] Time

Please specify how many months you expect your project to last

7

[Proposal Summary] Translation Information

Please indicate if your proposal has been auto-translated

No

Original Language

en

[Proposal Summary] Problem Statement

What is the problem you want to solve?

Indonesia accounts for 38% of APAC employment fraud. Meanwhile, the hiring cycle is paralyzed by repetitive screening interviews, extending the average time-to-hire to 49 days per role.

[Proposal Summary] Supporting Documentation

Supporting links

• https://docs.google.com/spreadsheets/d/1iOZmv8QHvhaWK3Tr29HR9z8g5pGNsTzm/edit?usp=sharing&ouid=114783060178863143479&rtpof=true&sd=true
,
• https://retrai.sumbu.xyz
,
• https://drive.google.com/file/d/1O9birJ7AZW1RwI-uZalX1PXrxRdezm_t/view?usp=sharing

[Proposal Summary] Project Dependencies

Does your project have any dependencies on other organizations, technical or otherwise?

No

Describe any dependencies or write 'No dependencies'

No dependencies

[Proposal Summary] Project Open Source

Will your project's outputs be fully open source?

Yes

Please provide details on the intellectual property (IP) status of your project outputs, including whether they will be released as open source or retained under another licence.

License: Hybrid (Apache 2.0 Core / Proprietary AI) IP Strategy: We prioritize ecosystem trust while maintaining business sustainability:

Open Source Trust Layer: The Cardano Smart Contracts (Registry) will be fully open-source (Apache 2.0) and published on GitHub.

Commercial Product Layer: The AI Analysis Engine (Prompt Engineering & "Deep Screen" Logic) will remain proprietary. This protects our competitive advantage and prevents bad actors from reverse-engineering the screening criteria.

[Theme Selection] Theme

Please choose the most relevant theme and tag related to the outcomes of your proposal

Business Services

[Campaign Category] Category Questions

Describe what makes your idea innovative compared to what has been previously launched in the market (whether by you or others).

Existing platforms (JobStreet, LinkedIn) and standalone AI tools fail to solve the fragmentation and safety issues Indonesia's recruitment market.

The Current Market Failure:

1. The Cost-Safety Dilemma: Legitimate SMEs in Indonesia often lack the budget for premium platforms like LinkedIn or JobStreet. They are forced to use low-barrier platforms like Facebook, Instagram, Glints, or KitaLulus. However, these platforms are infested with scammers because they lack rigorous identity verification. Legitimate companies end up in the same pond as fraudsters, losing credibility
2. The Manual Screening Bottleneck: To assess soft skills and culture fit, recruiters currently have to call candidates one by one. This is unscalable. While AI tools like Flowmingo exist, they are ineffective plugins that add friction. Candidates hate downloading a new app just for one interview, and recruiters hate managing a separate dashboard.
3. The Repetition Fatigue: Candidates are forced to repeat the same introduction interview for every single job application, wasting hours on redundant labor.

Retrai leverages Cardano’s unique architecture to solve this with a safe, efficient, and affordable ecosystem:

1. The Crowd-Sourced Safety Innovation

Retrai breaks the cost vs. safety trade-off. We offer affordable job ads (accessible to SMEs) but secured by a decentralized anti-scam registry that keeps the platform clean.

1. The Cardano Edge (CIP-68)

We utilize the CIP-68 (Datum Metadata Standard) to create living reputation identities. Unlike static NFTs, CIP-68 allows our moderators to verify user flags and update an employer's status (e.g., appending a Risk: Misleading tag) in a reference UTxO instantly and cheaply.

Why it Matters: This creates a censorship-resistant bad actor registry. It solves the $422 million scam crisis by creating a permanent, shared blocklist. Scammers who are banned cannot simply make a new email to reset their reputation because their on-chain identity is flagged. This gives cheap job ads the high security of a premium platform.

1. The One-Time Deep Screen Innovation

Retrai is the first to natively integrate a human-calibrated multimodal AI that solves the efficiency crisis for both sides of the market.

• For Recruiters (Zero Manual Screening): Instead of calling 50 candidates to check English fluency or attitude, recruiters receive a dashboard where these traits are already probed. Our AI asks hyper-personalized follow-up questions (e.g., "Why is there a gap in your resume in 2023?") that static video tools miss. This replaces the first phone screen entirely, saving a lot of hiring time. Crucially, our AI does not reject nor score candidates. It simply automates the repetitive task of the first screening process. The AI probes for details, compiles the data, and presents it to the recruiter. The recruiter retains 100% autonomy to watch the clips and make the final shortlist decision, saving them hours of call time without surrendering control to an algorithm.
• For Candidates: Unlike other platforms where candidates interview per job, Retrai introduces the one-and-done interview. A candidate completes the deep AI screening once during profile creation. This verified video profile becomes their asset. They can apply to as many jobs as they want instantly, and all recruiters can view the same high-quality, deep-dive interview. This massive convenience drives candidate adoption.

1. Risk Management Strategy

We have designed specific protocols to mitigate the risks inherent in reputation and AI systems:

• Mitigating Mob Rule: To prevent malicious candidates from brigading innocent companies, our CIP-68 Registry is permissioned at the moderator level for the MVP. User flags are treated as signals, but only a human moderator (checking evidence/screenshots) can sign the on-chain transaction to apply a tag. This ensures due process while maintaining ledger transparency.

• Mitigating AI Bias: We explicitly address the risk of AI hallucination by strictly limiting the AI's role to data collection, not decision making. The shortlist/reject button is exclusively controlled by the human recruiter.

Describe what your prototype or MVP will demonstrate, and where it can be accessed.

At the end of this 6-month project, we will deliver a fully functional Web Application connected to the Cardano Preprod Testnet, powered by Aiken Smart Contracts. The MVP will demonstrate the complete safety and efficiency loop:

1. For Candidates:

• Invisible Web3 Onboarding: A user signs up using Email/Google. The system uses Wallet Abstraction to automatically provision a non-custodial Cardano wallet in the background, removing the need for seed phrases while enabling on-chain interaction.
• One-Click Application: Once the candidate completes their AI-led screening interview, their profile is complete. They can apply to multiple jobs instantly with a single click, signing the application transaction via their abstracted wallet.
• Flagging Logic: Candidates can flag a company (Legit/Fraud/Misleading). The system submits this report to the moderation queue. Once validated by a moderator, the status is updated on the CIP-68 Registry.
• The "Shield" UI: Users viewing a job listing will see a prominent "COMMUNITY WARNING" badge if the employer's on-chain identity holds a Risk Tag, or a "VERIFIED LEGIT" badge if they passed verification.
• The "Public Link": Candidates receive a shareable Web2 URL for their verified video profile to use on external CVs.

1. For Employers:

• Integrated Dashboard: A recruiter can post a job and view all applicants in a single list. Instead of scheduling phone screens, they simply click a candidate's profile to watch the pre-recorded, AI-probed interview.
• Manual Shortlisting: The dashboard allows recruiters to explicitly shortlist or eeject candidates based on this deep data, maintaining full human control over the hiring decision.

Access & Transparency: The MVP will be accessible via a public beta link (e.g., beta.retrai.id) and the source code for the Aiken smart contracts (Registry Logic) will be published on our GitHub repository to ensure full community auditability.

Describe realistic measures of success, ideally with on-chain metrics.

We define success through a balanced scorecard of Trust Activity, Adoption, Efficiency, and Technical Delivery. Our goal is to prove that Cardano can serve as the invisible trust layer for a high-volume labor market.

1. Trust and Reputation Metrics

This is the core utility of our CIP-68 implementation. Success means the community is actively using the blockchain to filter bad actors.

• 50+ On-Chain Reputation Events: We aim to log at least 50 verified reputation actions on the Cardano blockchain via our CIP-68 Registry Contract during the 7-month pilot. This metric is critical because it proves the system is not just a database, but a living, decentralized ledger. Validated community reports flagging entities as legit/fraud/misleading are all counted. Each flag is an immutable transaction that protects future candidates.
• Why this matters: Achieving 50+ events demonstrates that this mechanism is functional and that users trust the system enough to engage with it.

1. Adoption Metrics

We focus on invisible onboarding, which is bringing non-crypto natives onto Cardano without friction.

• 500+ Gen-Z Wallets Onboarded: We aim to onboard at least 500 job seekers on Retrai. By implementing Wallet Abstraction, every student who signs up (via Google/Email) is automatically provisioned with a non-custodial Cardano wallet in the background.
• Success Indicator: 500+ distinct wallet addresses interacting with our smart contract (signing reputation reports) without users needing to manage seed phrases.
• 5 Active Job Listings: We aim to have at least 5 legitimate job posts created by verified beta partners. This proves we can attract the supply side (employers) of the marketplace, validating the B2B value proposition.

1. Efficiency Metrics

This measures the business value we deliver to employers, which is the key to long-term sustainability.

• 1,000+ Minutes of AI Screening: We aim to process at least 1,000 minutes of candidate interview video through our Gemini 3 engine. This metric directly validates the utility of the deep screen feature. It means candidates are actually interested and engaged with the core feature.
• 70% Time Reduction: We will conduct feedback surveys with beta recruiters to confirm that the AI-led screening feature saved them the need to conduct manual screening calls. Success is defined as a 70% reported reduction in time-to-shortlist compared to their previous manual process.

1. Technical Delivery

• Open Source Release: Publication of the Aiken Smart Contracts (Registry logic) on GitHub, fully documented for community auditing. Success is defined by the code passing internal security review and being forkable by other developers.
• Mainnet Readiness: While the pilot runs on Preprod for safety, success includes a fully tested deployment script ready for Mainnet launch, ensuring no technical debt blocks the commercial rollout.

[Your Project and Solution] Solution

Please describe your proposed solution and how it addresses the problem

The Indonesian recruitment landscape is paralyzed by two opposing forces: predatory fraud and operational inefficiency.

1. The predatory crisis ($422M loss)

Indonesia accounts for a staggering 38% of all employment fraud in the Asia-Pacific region. Legitimate SMEs are caught in a cost vs. safety dilemma: they cannot afford the high fees of premium platforms like JobStreet ($300+/ad), so they are forced to use low-barrier platforms like Facebook, Instagram, Glints, or KitaLulus. However, these platforms are infested with scammers because they lack immutable identity verification. Predatory "futures brokerage" firms routinely exploit fresh graduates by disguising commission-only, high-risk sales jobs as salaried "management trainee" roles. It is a financial trap costing the Indonesian economy Rp 7 Trillion ($422 million) annually. Scammers banned today simply create a new email and post a fake admin job tomorrow because there is no permanent "bad actor registry."

1. The inefficiency crisis

Legitimate employers are drowning in noise. Our preliminary research of 18 Indonesian recruiters (2024) revealed a critical bottleneck: recruiters rated "communication" and "attitude" as the hardest traits to assess via CV (difficulty score: 4.5/5).

• The manual bottleneck: because text resumes fail to capture these soft skills, HR managers are forced to manually phone-screen hundreds of applicants and ask basic screening questions like "why is there a gap in your resume?"
• The tooling failure: existing video tools (like Flowmingo) are ineffective "plugins" that add friction. Candidates hate downloading new apps for every job, and recruiters hate managing separate dashboards. This inefficiency extends the average time-to-hire to 49 days, stifling business growth.

The Solution:

Retrai is a dual-sided marketplace built on Cardano that unifies the job board, the AI screener, and the trust registry into a single, secure ecosystem. We replace the "black box" of Web2 hiring with a transparent, on-chain reputation system.

1. Core component 1: community-curated Risk Registry for the Indonesian labor market, powered by Cardano CIP-68.

• Crowdsourced Vigilance: Candidates are the first line of defense. When a candidate encounters a company on Retrai, they can flag the listing with specific reputation tags: "Legit" (Verified), "Fraud" (Scam), or "Misleading" (Bait-and-Switch).
• Moderator Verification: To prevent "mob rule," user flags are treated as signals. Our moderation team reviews the evidence (screenshots/chats) and validates the claim.
• Immutable Consequences (CIP-68): Validated reports result in a permanent "Risk Tag" being appended to the employer’s on-chain identity. Because this metadata lives on the Cardano blockchain via the CIP-68 Datum Standard, it cannot be bribed away, deleted by a corrupt admin, or hidden. It serves as a permanent, public warning label that protects future applicants for years to come.

1. Core component 2: AI-led deep screening to solve the 49-day bottleneck, we introduce an autonomous AI interviewer that replaces the manual phone screen.

• One-and-done profile: unlike other platforms where candidates interview per job, Retrai introduces the "\profile-first interview. Candidates complete the deep AI screening once during onboarding. This verified video profile becomes their asset. They can apply to as many jobs as they want instantly, and every recruiter sees the same high-quality interview.
• Psychologically validated probing: unlike generic tools, Retrai’s AI is trained on expert-designed interview frameworks. It reads the candidate's CV and generates hyper-personalized questions. If a candidate gives a vague answer (e.g., "I am a hard worker"), the AI triggers a context-aware follow-up: "can you describe a specific situation where that hard work solved a crisis?"
• Zero-bias assistance: the AI does not score or judge the candidate. It acts purely as a structured data gatherer. The recruiter retains 100% autonomy to watch the clips and make the "hire/no hire" decision based on their own cultural fit standards.

1. Core component 3: frictionless adoption (wallet abstraction & growth) we bridge the gap between Web2 usability and Web3 security.

• Invisible Web3 (wallet abstraction): we recognize that students will not use a platform that requires managing seed phrases. Retrai implements wallet abstraction. When a candidate signs up with Google/Email, a non-custodial Cardano wallet is automatically provisioned in the background. This allows them to interact with the blockchain (signing reputation reports, holding credentials) without ever realizing they are using crypto.
• The public link: candidates receive a public, shareable URL for their verified interview. They can paste this link onto their standard PDF CVs when applying to companies outside of Retrai. When a Web2 recruiter clicks the link, they see the "verified on Cardano" badge, driving organic traffic back to our ecosystem.
• B2B "credit" model: employers purchase "Retrai credits" (internal platform points) to post jobs. This generates sustainable revenue to fund the "anti-scam" operations.

Why Cardano?

Retrai is economically viable only on Cardano due to three specific architectural advantages that solve problems Web2 and EVM chains cannot:

1. Decoupled state via CIP-68 (the "living" registry) standard NFTs (ERC-721) are static; updating metadata requires burning the token, which is expensive and destroys history.

• The architecture: we use the CIP-68 (datum metadata standard). This separates the user token (held by the employer) from the reference UTxO (holding the reputation data).
• The capability: our registry smart contract can update the "risk status" (e.g., adding a warning: fraud tag) in the reference UTxO without needing the employer's private key. This allows for a dynamic, evolving reputation system that is censorship-resistant yet updatable by community consensus.

1. Native assets & eUTxO (cost-efficiency) on Ethereum, minting reputation signals or credentials requires executing complex smart contract code for every item, costing significant gas. The advantage: Cardano treats tokens as native assets. We can mint thousands of "reputation signals" in a single transaction bundle for negligible fees (~0.17 ADA min-ADA). This makes high-volume B2B tokenization sustainable in a developing market like Indonesia.

2. High-assurance security (Aiken)

• The advantage: we write our registry logic in Aiken, leveraging Cardano’s eUTxO model. This ensures that transaction costs are predictable and execution is deterministic, meaning a "scam flag" transaction will never fail mid-execution and waste fees, a critical reliability feature for a business-critical application.

[Your Project and Solution] Impact

Please define the positive impact your project will have on the wider Cardano community

Retrai moves the Cardano narrative beyond defi speculation to build Critical Business Safety Infrastructure. By addressing the $422 million employment fraud crisis in Indonesia, we position Cardano as the only blockchain capable of solving a massive, real-world trust deficit that Web2 platforms have failed to fix for a decade.

1. Generating High-Quality real-world Transaction Volume

Most RealFi projects suffer from low on-chain activity because users only interact once (e.g., minting an ID). Retrai is architected to generate continuous, recurring transaction volume through our feedback loop.

• The Mechanism: Every candidate who applies for a job enters a verification cycle. 7 days after applying, they are prompted to verify the legitimacy of the employer using three specific flags: legit, fraud, or misleading.
• The Volume: With a conservative target of 500 active beta users, we project hundreds of verified reputation signals per month. Each flag triggers a state update in the CIP-68 Registry Contract, flooding the Cardano network with meaningful, non-financial data. This proves that Cardano can handle high-throughput social signal processing, a key requirement for mass adoption.

1. Commercial Sustainability (B2B Revenue Model)

We are not building a project that will die when the Catalyst grant runs out. Retrai addresses a $6 Billion HR Tech Market with a clear path to profitability.

• The Value Proposition: Employers are desperate for speed. Our preliminary research of 18 Indonesian Recruiters confirmed that communication and attitude are the hardest traits to assess, causing a 49-day hiring cycle. Employers are willing to pay for our AI deep screen because it reduces this cycle by 70%.
• The credit Economy: We utilize a Dynamic Credit Model. Employers purchase retrai credits (internal platform points) using Fiat or ADA to post jobs and unlock AI screening. This generates immediate cash flow to fund server costs (Gemini 3 API, DigitalOcean) and scam hunt bounties, ensuring the platform is self-sustaining from Year 1.

1. Corporate Onboarding (The trojan horse Strategy)

Retrai acts as a powerful trojan horse for onboarding Indonesian SMEs (Small and Medium Enterprises) onto Cardano.

• The Frictionless Entry: Legitimate companies join Retrai not because they care about decentralization, but because they want cheaper ads and faster hiring process.
• The Implicit Onboarding: To distinguish themselves from scammers and earn the verified legit badge, companies must interact with our on-chain registry (abstracted via a seamless UI). This implicitly onboards hundreds of Indonesian businesses onto the blockchain, creating a network of Verified Business Identities (KYB).
• Future Utility: Once these companies have a Cardano-based identity, they become reachable by other B2B dApps in the ecosystem (e.g., Decentralized Payroll, Invoice Factoring), creating a network effect that lifts the entire Cardano economy in Indonesia.

1. Mass Adoption via Wallet Abstraction

We solve the biggest barrier to Web3 adoption: User Friction.

• Invisible Onboarding: Through Wallet Abstraction, we provision non-custodial Cardano wallets for students automatically when they sign up via Email/Google. They do not need to understand seed phrases to participate in the Reputation Registry.
• The Result: We are creating 500+ active wallet holders who interact with the blockchain daily to manage their careers. This creates a pipeline of educated, income-earning users who are ready to explore the wider Cardano DeFi ecosystem once they are comfortable with their career wallet.

1. Open Source Contribution (The trust primitive)

We are not just building a proprietary app; we are contributing a foundational primitive to the Cardano developer community.

• The Deliverable: We will open-source our Aiken-based reputation registry contracts under an Apache 2.0 license.
• The Ripple Effect: This registry logic is universal. It can be forked by a Freelance DAO to track bad clients, by a P2P Marketplace to track scammers, or by a Lending Protocol to track defaulters. By solving the reputation problem for recruitment, we provide the code blocks for any other project that needs a shared blacklist, accelerating development for future builders across the ecosystem.

1. Social Impact: Solving a Verified $422 Million Crisis (RealFi)

The Indonesian labor market is currently a low-trust environment, which stifles economic growth.

• The Economic Cost: According to the Financial Services Authority (OJK), digital fraud and job scams cost the Indonesian economy an estimated Rp 7 Trillion ($422 million) between late 2024 and 2025 (Tech in Asia, 2025). fake job offers specifically accounted for Rp 656 billion ($39.6 million) in losses. These aren't just numbers; they represent fresh graduates losing their family savings to admin fee scams or being trafficked by fake overseas recruitment agencies.
• The Cardano Solution: Retrai provides a Public Good: an immutable, censorship-resistant bad actor registry. By enabling candidates to flag predatory actors on-chain via CIP-68, we create a permanent record that cannot be bribed away or deleted by corrupt admins. This protects the most vulnerable demographic (Gen Z graduates) and demonstrates Cardano’s mission of identity protection to the Indonesian government (Ministry of Manpower) and university partners.

[Your Project and Solution] Capabilities & Feasibility

What is your capability to deliver your project with high levels of trust and accountability? How do you intend to validate if your approach is feasible?

Our team combines high-velocity technical execution with deep domain expertise in psychology and HR. Every member is fully committed, dedicating at least 25 hours/week to ensure Retrai ships on schedule.

1. Giga H. A. Adkhy (lead architect & project manager)

Role: blockchain architecture & smart contract security.

Giga is a proven Catalyst executor. He served as the project manager for the successfully delivered Fund 13 "Indonesia Car-Dano" project. In that role, he led the development of a blockchain-based vehicle inspection registry, proving his ability to manage Treasury funds responsibly, hit milestones on time, and ship real-world dApps. This prior success eliminates the "delivery risk" often associated with new proposers.

Technical expertise: Giga designs the CIP-68 metadata standards and Aiken smart contracts that power our registry. His background in building UGM-AICare (AI mental health platform) ensures he knows how to securely integrate AI agents with blockchain logic.

GitHub: https://github.com/gigahidjrikaaa

LinkedIn: https://www.linkedin.com/in/gigahidjrikaaa/

1. Virna Amrita (product manager & domain lead)

Role: product strategy, AI calibration, & scam definitions.

Virna is a faculty member in psychology at Universitas Gadjah Mada. She earned a gold medal in PKM-RSH 2023 for her research on "PsyBot: WhatsApp-based psychological first aid." This proves that she has unique expertise in training LLMs and is very relevant to this project. She calibrates our AI "interviewer persona" to ensure it mimics a senior recruiter’s intuition rather than a robotic script. She also manages the scam definitions database, ensuring our flagging system aligns with Indonesian labor laws.

GitHub: https://github.com/virnamrita

LinkedIn: https://www.linkedin.com/in/virna-amrita-a13463261/

Research Gate: https://www.researchgate.net/profile/Virna-Amrita

1. Deren Tanaphan (Backend Engineer)

Role: core systems engineer (API & database)

Track record: lead developer for Quest.io, an object-oriented platform focused on SDG-8 (decent work). Deren specializes in building high-concurrency backend systems. He ensures that when 1,000 students apply simultaneously during a "job fair," our API handles the load without crashing.

GitHub: https://github.com/Lev1reG

LinkedIn: https://www.linkedin.com/in/derentanaphan/

Website: https://derentanaphan.web.id

1. Benaya Imanuela (Frontend Engineer)

Role: UI/UX implementation & wallet integration

Track record: developer at Gamaforce UGM and Arachnova.id. He built the Tekmart frontend using Vite. Benaya ensures the candidate dashboard hides the complexity of blockchain (wallets, transactions) behind a beautiful, intuitive interface, which is crucial for onboarding non-crypto natives.

GitHub: https://github.com/eben4ya

LinkedIn: https://www.linkedin.com/in/benaya-imanuela/

Website: https://eben.biz.id

• Feasibility strategy

We minimize delivery risk by adhering to a strategy of code reuse, data readiness, and hybrid development.

A. Technical validation (the "Fund 13" blueprint)

We are not inventing the "trust registry" concept from scratch. We are refactoring proven architecture.

Impact: this reduces the backend R&D risk by approximately 40%, allowing us to focus our energy on the AI integration and frontend UX.

B. Operational validation (agile process)

1. Hybrid open-source model: to ensure business sustainability, we keep the AI prompt engineering proprietary (protecting us from copycats) while fully open-sourcing the trust layer (Aiken smart contracts) under Apache 2.0. This balances community transparency with commercial defensibility.
2. Beta partner network: we have direct access to the student bodies of UGM for our beta launch. This guarantees we will hit our target of 500 active profiles without needing an expensive marketing budget.

• Financial accountability

We treat the Catalyst treasury with extreme respect.

A. Budget efficiency: by leveraging the lower cost of living in Yogyakarta, we deliver a 4-person team for 45,000 ADA, a cost that would barely cover one month of a single developer's salary in the US/EU.

B. Milestone discipline: as demonstrated in Fund 13, we adhere to strict milestone reporting standards (SoM/PoA), ensuring funds are only released upon verifiable proof of delivery.

[Milestones] Project Milestones

Milestone Title

Foundations: Trust Registry Architecture & AI Persona Design

Milestone Outputs

• Design System (Frontend): Final High-Fidelity Figma prototypes for the "Candidate Safety Dashboard" (Reporting Flow), "Recruiter Efficiency View" (Video Player), and "Job Card" with Risk Badges.
• Smart Contract Architecture: Technical Specification Document defining the CIP-68 metadata structure for the "Risk Tags" (Legit/Fraud/Misleading) and the access control logic for moderators.
• AI Logic Design: A finalized "Interviewer Persona" script (System Prompts) designed by our Domain Lead to ensure psychologically valid probing questions.

Acceptance Criteria

• The Figma designs must cover the end-to-end user journey: Connect Wallet -> Apply to a Job -> Flag a Company -> View Verified/Warning Badge.
• The Technical Spec must be reviewed and approved by the Lead Architect, clearly documenting the Aiken validator logic.
• The AI Script must be demonstrated in a local text-based environment, successfully triggering a "Follow-Up" question when a generic answer is detected.

Evidence of Completion

• Figma Link: A public URL to the finalized UI/UX design file.
• PDF Spec: A "Retrai Architecture v1.0" document uploaded to the repository.
• Demo Video: A screen recording showing the AI prompt logic working in a local environment.

Delivery Month

2

Cost

9000

Progress

20 %

Milestone Title

Smart Contract & Backend API Construction

Milestone Outputs

• Smart Contracts (Trust Layer): Development of the Aiken Registry Contract that handles permissioned updates for "Risk Tags" (e.g., flagging a scammer).
• Backend API (Infrastructure): Implementation of the high-concurrency REST/GraphQL API that receives user reports, stores evidence (screenshots), and queues them for moderation.
• Reputation Database: Setup of the off-chain indexer to track "Trust Points" for users who submit valid reports.

Acceptance Criteria

• The Aiken Smart Contract must compile successfully and pass unit tests for "Minting" and "Updating" reputation tags on a local emulator.
• The Backend API must successfully receive a "Report" payload and store it in the database.
• Integration Test: A dummy report submitted to the API successfully triggers a status update in the local test database.

Evidence of Completion

• GitHub Repo: Link to the public repository containing the Aiken source code.
• API Docs: Link to the Swagger/OpenAPI documentation.
• Test Report: A screenshot of backend logs showing successful report submission.

Delivery Month

4

Cost

13500

Progress

50 %

Milestone Title

Frontend, AI Integration & Feedback Loop

Milestone Outputs

• Frontend Integration: Full React/Vite web interface integrated with the Backend API.
• Wallet Integration: Implementation of Wallet Abstraction (Email Login) for frictionless onboarding. Note: Standard Wallet Connect (Vespr/Eternl) will be implemented as a fallback to ensure delivery.
• AI Voice Interface: Integration of the Gemini 3 Multimodal API into the frontend, allowing users to record real voice answers.
• "Feedback Loop" Logic: Scheduler implementation to trigger "Weekly Legitimacy Check" notifications.

Acceptance Criteria

• A user can connect a wallet (via Email Login OR Standard Wallet) and view the dashboard.
• A user can complete a voice interview session, and the system returns a text transcript.
• A "Flagged" job post correctly displays the Red Warning Badge on the live frontend.
• The scheduler successfully triggers a test notification to a user account.

Evidence of Completion

• Live Beta Link: URL to the deployed web app (e.g., Vercel/Netlify) running on Preprod.
• Video Walkthrough: A full video showing the flow: Connecting wallet -> Applying -> Flagging a Job -> Taking an AI Interview.
• Code Commit: GitHub link showing the frontend integration code.

Delivery Month

6

Cost

13500

Progress

80 %

Milestone Title

Launch: Public Beta, Validation & Project Close-out

Milestone Outputs

• Beta Execution: Launching the platform to initial users (UGM students) for live testing
• Population of Registry: Manual population of the registry with initial known scams/legit companies
• Open Source Release: Final documentation (README) and cleanup of the GitHub repository
• Project Close-out: Submission of the Final Report and Final Impact Video to Catalyst

Acceptance Criteria

• Trust: System logs show at least 50 Reputation Flags (legit/fraud/misleading) processed during beta.
• Adoption: At least 5 active job listings are live on the platform
• Efficiency: Beta survey results collected from recruiters
• Tech: The GitHub repository must be public with a "Mainnet Deployment Script" ready

Evidence of Completion

• Impact Report: A PDF summarizing the beta metrics (Wallets Onboarded, Flags Logged, AI Minutes Processed).
• On-Chain Data: A list of Transaction IDs (TxIDs) proving the registry updates on Preprod/Mainnet.
• Survey Report: A PDF of the survey report, censoring the participants of the survey
• Close-out Video: A video demonstrating the platform's features and real-world usage.

Delivery Month

7

Cost

9000

Progress

100 %

[Final Pitch] Budget & Costs

Please provide a cost breakdown of the proposed work and resources

We have engineered an infrastructure-first budget. By capping human capital expenses at just 40%, we are able to direct 60% of the funding into enterprise-grade servers, AI processing power, and security auditing. This ensures Retrai launches as a high-performance, scalable platform from Day 1.

1. Human Capital (40% - 18,000 ADA)

Fixed technical stipends for the 4-person squad for the 7-month build.

• Giga H. A. Adkhy (Lead Architect): 5,000 ADA total.

Deliverables: Aiken Smart Contract Security, System Architecture, Milestone Reporting.

• Virna Amrita (Product Manager): 5,000 ADA total.

Deliverables: AI Persona Calibration, Scam Database Curation, Product Roadmap.

• Deren Tanaphan (Backend Lead): 4,000 ADA total.

Deliverables: API Scalability, Database Sharding, High-Concurrency Architecture.

• Benaya Imanuela (Frontend Lead): 4,000 ADA total.

Deliverables: UI/UX Implementation, Wallet Connect Integration, Mobile Responsiveness.

1. Operational Expenses (10% - 4,500 ADA)

Essential costs to facilitate intense collaboration and partner onboarding.

• Meeting & Co-working Costs: 2,500 ADA
• Usage: Weekly sprint planning sessions and intensive coding weekends for the 4-person team.
• Transportation & Logistics: 2,000 ADA.
• Usage: Travel costs for on-site meetings with university partners (UGM Career Center) and recruitment firms to secure beta users.

1. High-Performance AI Compute (25% - 11,250 ADA)

Allocating massive computational resources to the Gemini 3 Multimodal Engine.

• Gemini 3 API Tokens: 11,250 ADA.
• Tech: Google Gemini 3 Pro (Native Video/Audio Support).
• Usage: Budgeted for ~7,500 hours of high-fidelity video processing.
• Why this matters: Video input tokens are computationally expensive. This large buffer ensures we can handle a massive spike in user traffic (e.g., during university job fairs) without hitting API rate limits or degrading the deep screen experience.

1. Cloud Infrastructure and Security (25% - 11,250 ADA)

Building a banking-grade environment to protect user reputation data.

• Scalable Compute (Hostinger): 4,000 ADA.

• Item: VPS KVM 8 (8 vCPU / 32GB RAM) for dedicated audio/video transcoding microservices. We selected Hostinger for its superior price-to-performance ratio in the SEA region compared to AWS.

• Item: Managed Database with automated failover and daily backups.

• Web3 Infrastructure: 3,000 ADA

• Item: Blockfrost Enterprise Plan for high-frequency on-chain reputation queries (checking CIP-68 tags in real-time)

• Item: Pinata Dedicated Gateway for low-latency retrieval of decentralized metadata

• Security & DevOps Suite: 4,250 ADA

• WAF: Cloudflare Pro for unmetered DDoS mitigation and bot protection

• Monitoring: Sentry Team Plan for real-time backend error tracking and performance tracing

• CI/CD: GitHub Team runners for automated testing pipelines

[Final Pitch] Value for Money

How does the cost of the project represent value for the Cardano ecosystem?

We are requesting 45,000 ADA to solve a verified multi-million dollar crisis. This proposal offers exceptional value by leveraging geographic labor arbitrage and targeting a high-growth market sector.

1. Solving a $422 Million Problem with Micro-Funding

• The Cost of the Problem: According to the Financial Services Authority (OJK), digital fraud and job scams cost the Indonesian economy an estimated Rp 7 Trillion ($422 million) between late 2024 and 2025 (Tech in Asia, 2025). Specifically, "Fake Job Offers" accounted for Rp 656 billion ($39.6 million) in losses (Jakarta Globe, 2025).
• The ROI: For a grant of just 45,000 ADA, Retrai builds a permanent, decentralized "Shield" against this massive financial drain. If our "Risk-Flagging Registry" prevents just 0.1% of these job scams, we will have saved the ecosystem over $39,000—paying for the entire project in its first year of operation.

1. Strategic Entry into a $6 Billion Growth Market

• The Market: The Indonesian HR Technology market was valued at $6.01 billion in 2024 and is continuing to grow as digitalization accelerates (Business Indonesia, 2025).
• The Opportunity: By funding this pilot, Catalyst is buying an early stake in the dominant HR tech infrastructure of Indonesia. We are not building a "niche dApp"; we are building the "Trust Layer" for a workforce of 153 million people that is rapidly digitizing (Rockbird Media, 2025).

1. Long-Term Adoption: Onboarding the "Class of 2030"

• The Pipeline: Data from the Ministry of Education confirms that Indonesia produces approximately 1.85 million university graduates annually (Tempo.co, 2023). These are young, tech-savvy Gen-Z professionals who are currently unbanked or under-banked.
• The Strategy: Retrai does not market "Crypto Trading" to them. We market "Career Safety." By integrating Cardano wallets (e.g., Vespr) as the default "Career Passport" for our partner university students, we implicitly onboard thousands of future developers and professionals into the ecosystem.
• Lifetime Value: We are acquiring users at a Customer Acquisition Cost (CAC) of ~$4.50 (based on our user incentives budget), which is significantly lower than the fintech industry average ($20-$50).

1. Geographic Efficiency (Local Team)

Cost Efficiency: Our budget leverages the lower cost of living in Yogyakarta. We are delivering a 4-person R&D squad (Full Stack, Blockchain, AI, Psychology) for 7 months at a cost that would barely cover one month of a single Silicon Valley engineer's salary. This ensures 100% of the Treasury’s funds go toward code and product, not inflated overheads.

[Self-Assessment] Self-Assessment Checklist

I confirm that evidence of prior research, whitepaper, design, or proof-of-concept is provided.

Yes

I confirm that the proposal includes ecosystem research and uses the findings to either (a) justify its uniqueness over existing solutions or (b) demonstrate the value of its novel approach.

Yes

I confirm that the proposal demonstrates technical capability via verifiable in-house talent or a confirmed development partner (GitHub, LinkedIn, portfolio, etc.)

Yes

I confirm that the proposer and all team members are in good standing with prior Catalyst projects.

Yes

I confirm that the proposal clearly defines the problem and the value of the on-chain utility.

Yes

I confirm that the primary goal of the proposal is a working prototype deployed on at least a Cardano testnet.

Yes

I confirm that the proposal outlines a credible and clear technical plan and architecture.

Yes

I confirm that the budget and timeline (≤ 12 months) are realistic for the proposed work.

Yes

I confirm that the proposal includes a community engagement and feedback plan to amplify prototype adoption with the Cardano ecosystem.

Yes

I confirm that the budget is for future development only; excludes retroactive funding, incentives, giveaways, re-granting, or sub-treasuries.

Yes

[Required Acknowledgements] Consent & Confirmation

I Agree

Yes