📈

Investment Memo

Looking to invest in Atadia? Read this.

Summary

  • For Web3 to get closer to mass adoption, one of the most important data infra is still missing: insightful & performant on-chain user profile data.
  • While Big Data and Advanced Analytics enabled tremendous value in Web2 and IRL, most of Web3 still rely little of it. This limits its true potential and slows down the emergence of new, more powerful (or more mass-familiar) products and UX.
  • 3 highest upside opportunities we think that Web3 can do even better with on-chain user profile data are:
    • Customer Profiling => instant CRM, fair data monetization, soulbound identity
    • Counter-party risk mngmt => reputation, new forms of credit, safety surveillance
    • Digital advertising => faster product & community discoverability, Web3 data economy
  • To enable these opportunities, we spent the past few months building a block-to-profile data pipeline on Solana. There are 500+ attributes and counting.
  • Credit Risk is our BIGGEST use-case. Afawk, we are the only Web3 project that has underwritten a significant amount of zero-collateral, zero-KYC loans purely based on ML stack trained with on-chain data. Default rates are now comparable or lower than most over-collateralized loans.
  • Our core team is a small IRL data science solutions company that pivoted to Web3 in 2022. Our technical expertise is in Machine Learning and our functional expertise is in alternative credit, market surveillance, and retail analytics.
  • We bootstrapped our 1st NFT Collection mint fund to date and are looking to raise additional funds to continue to build on top of the foundation we've already built.
Underlying assumptions:
  1. 1.
    Web3 will eventually take off and go beyond trading coins and NFTs
  2. 2.
    We will see demand for credit-based or trust-based DeFi products on Web3
  3. 3.
    There is a need for algorithmic services on Web3
  4. 4.
    All the 3 upside opportunities are actually attainable despite being super challenging

Introduction

Web3 has the power to disrupt how consumer data are collected, distributed, and commercialized.
Permissionless, decentralized networks and public transaction records help lay the foundation that can enable the next chapter of advanced analytics. This is the tech that powered the Platform Revolution across industries like digital advertising, e-commerce, and fintech. Coming into Web3 from data science & business analytics space, we see 3 core advantages here
  1. 1.
    Distribution — because transactions are public, anonymized consumer data should be more readily available to anyone and not kept in the hands of a few
  2. 2.
    Integrity — because there is a single source of truth, data can be more easily verified and/or disputed if needed
  3. 3.
    Monetization — with native payment infrastructure already built in, rewarding data owners and those who add value is much easier to program and execute
These efficiency and equity improvements will eventually give rise to a far more superior analytics-enabled commercial infrastructure that power new products and services not seen anywhere before.
We foresee that any online business would be able to utilize user data even if it’s the first visit. Businesses will not need to solely rely on building their own user database or on a handful of data giants.
From the user point of view, consumers get customized experiences and offers in more venues, but now with KYC and privacy settings on their own terms. Plus, they automatically receive rewards any time someone commercializes their data.
This is both a multi-trillion dollar opportunity and a major disruption waiting to happen.

Problem

“Better Big Data without Big Tech” is much easier said than done. In fact, it's beyond chewing glass.
To date, Web3 is still plagued with friction points that are hindering ecosystem growth. Barely any dApp or NFT community is using even basic historical on-chain profile data to enhance their value propositions or expand product offerings.
You may be wondering why that is the case when it seems like all the stars should have already aligned by design.🤔
The problem is that it’s too challenging and costly to ingest, parse, store, and serve on-chain consumer data in a performant way.
And so, Web3 consumer data often go severely underutilized while Web2 and IRL incumbents are maxing out on milking these data to cement their competitive edge. This results in limited product offerings, poor visibility & transparency, slow discoverability, and subpar risk management practices.
Let’s look at all the enabling pillars that support the new data economy and the gaps:
  • Access/Depth/Speed: how accessible are the data? just for devs? at what depth can you get them? do you need to wait hours for large data pulls?
  • Profile Enrichment: are the data enriched enough to be useful? Or are they just dressed-up gibberish?
  • Algo-ready: can we go beyond descriptive dashboards? how hard is it to build analytics algorithms?
  • Execution: can we also perform or automate on-chain txn based on these data?
Text
Access/Depth/Speed
Profile Enrichment
Algo-ready
Data-driven Execution
Node Providers/Dev Platforms
✅ mostly for devs
✅ sometimes
Explorers
✅ most still have poor context
Blockchain Analytics
Wallets
Web3 CRM
✅ sometimes
✅ sometimes

Solution

Atadia is building gmDATA, a platform that ingests on-chain txn data, turns them into profiles, and enables various data-driven utilities for dApps, communities, and everyday users.
Our goal is to make any tasks involving Web3 user profile data as easy as saying “gm.”
We started building gmDATA because there was absolutely zero infra that could support a complex MLOps project for on-chain credit scoring, our initial focus back in April 2022.
Text
Without Atadia 😢
With Atadia ⚫⚪
Access/Depth/Speed
  • Few allow access to parsed, indexed, and contextualized data
  • Even reputable providers sometimes deliver incomplete data or have tight rate limit 😢
  • It can take minutes, if not hours, to construct on-chain profiles at scale
  • Access profile data in seconds whether you're a dev or not
    • API
    • Data Soulbound Tokens
    • Dashboards (coming soon)
  • We triple check accuracy from various sources
Profile Enrichment
  • shallow level info such as NFT and token holding, volume, and txn value
  • Historical + recent
  • More than 500+ attributes already labeled and counting
  • Variety of levels (address, token, collection, and user level estimated from our algo)
Algo-ready
  • Zero to none
  • Data structured for MLOps
  • Purposed built profiles and scores that add value to projects and ecosystem as a whole
    • Web3 credit scores
    • Diamond hand score
    • Background check score
  • On-chain algorithmic systems
    • Atadia Wash Trade Alerts
    • More!
Data-driven Execution
  • Mostly manual or very tedious to do
  • Lots of errors
  • Profile data-driven token distribution as a service (coming soon)
  • Other advanced operations
    • Lending based on profiles

Credit: our biggest use-case

The highest upside use-case of Atadia's Web3 data capability is on-chain credit risk.
Since April 2022, we have built the capability to securely operate the Lending Lab, a dApp that underwrites zero-collateral loans purely based on on-chain profiles from the gmDATA platform. Repayment results get fed back into our ML model to retrain it.
From 3 phases of Lending Lab, we have originated approximately ~$300,000 USD worth of zero-collateral, zero-KYC short-term loans.
Our application basis default rates have come down from ~26% in Phase 1 to roughly 3-4% in Phase3. Volume basis default rates have come down from 38% in Phase 1 to around 3-4% in Phase 3.
Default rate can go up when we scale more adventurously
There are few Web3 Credit Risk players, but they either don't do credit scoring at all or are behind Atadia in both data science capability and underwriting.
To date, Atadia is the only zero-collateral lender on Web3 with this level of technical capability, tried-and-true lending, and credit scoring IP not shared with any VC.

Achievements so far

  • ~2000 members have minted soulbound tokens with gmDATA generated profiles
  • 5 profile metrics are dynamically tracked and updated
  • End-to-end, zero-collateral lending stack that has disbursed over $300,000 worth of SOL loans
  • ML-based credit scoring model trained from 3 phases of Lending Lab
  • 100+ projects, 8000+ WL spots, +23,000 applications
  • PoLo WL spot winners on avg. hold NFTs 2 weeks longer

Founder & Team

Puppet (@puppetatadia) is the founder of Atadia. He has 10+ years of Big Data and AI experience in various industries from telco, banking, capital market surveillance, and retail. His background is a PhD economist-turned-tech entrepreneur. He has recently published a book on how to improve the way we use data technologies based on his experience consulting for multi-billion dollar companies and propelling more evidence-based public policy in Thailand. The book (also named Atadia) would inspire him to create Atadia for real in March 2022.
He's a long-time member of MonkeDAO and a Bus Driver of Monkey Baby Business (MBB).
Atadia Core Team is from Puppet's IRL company. We are mostly data scientists, data engineers, economists, and devs.
See for team page for details.

Web3 Advisors

Atadia would not be here without the support from our advisors. 🤝
  1. 1.
  2. 2.
    RaviMonke
  3. 3.
    Branche
  4. 4.
    DegenTim
  5. 5.
    Gaius
  6. 6.
    HGE
  7. 7.
    Madhatt3r

Fundraise

We have been operating with ~$500K budget after we minted in March 2022.
We’re aiming to raise enough funds to finance product development, capital to scale Lending Lab, payroll with 2 more developers with at least 2 years of expected runway. If you would like to invest, don't miss our NeoAtadians Hybrid IDO!

Our priorities in 2023

  1. 1.
    increase gmDATA performance, cost-efficient scalability, and verifiability
  2. 2.
    scale Lending Lab in relations to $ATA with more types of products
  3. 3.
    lead effort to increase ecosystem safety on Solana
  4. 4.
    iterate data-driven services to get closer to short & medium term PMF
  5. 5.
    move closer to embracing open-source where we can
Last modified 1mo ago