With the amount of new subnets being added it can be hard to get up to date information across all subnets, so data may be slightly out of date from time to time
With the amount of new subnets being added it can be hard to get up to date information across all subnets, so data may be slightly out of date from time to time
The oneoneone Subnet is a decentralized AI data network built on the Bittensor blockchain. Its goal is to collect, validate and serve authentic user-generated content (reviews, posts, comments, etc.) from across the web for AI applications. Key features include:
oneoneone’s function is to turn “real conversations” (reviews/comments) into “intelligent data” in a censorship-resistant way, leveraging cryptoeconomic incentives.
The oneoneone Subnet is a decentralized AI data network built on the Bittensor blockchain. Its goal is to collect, validate and serve authentic user-generated content (reviews, posts, comments, etc.) from across the web for AI applications. Key features include:
oneoneone’s function is to turn “real conversations” (reviews/comments) into “intelligent data” in a censorship-resistant way, leveraging cryptoeconomic incentives.
OneOneOne operates with three participant roles:
Miners: Run scraping agents (Node.js + Apify) that gather and format content from web platforms. Each miner has an API key (e.g. Apify token) and a stake in the subnet, and they launch data-collection jobs in response to challenges.
Validators: Issue and evaluate tasks. Every ~20 minutes a validator creates a synthetic challenge (e.g. “scrape all reviews for a random restaurant on Google Maps”). They randomly select up to 50 miners to participate, then verify the returned data. Validators check speed, format and accuracy (e.g. spot-checking responses against live platform data) to enforce quality. Validators also batch the outputs into the structured dataset.
API Consumers: Clients access the cleaned data via the oneoneone.io API. They pay subscription fees for this data.
The data pipeline proceeds in steps:
Data Scoring Mechanism
To rank miner submissions, oneoneone uses a cryptoeconomic scoring formula. Each miner’s score S is computed as:
S = 0.5 × V + 0.3 × T + 0.2 × R
where:
This 50/30/20 weighting (Volume/Speed/Recency) is enforced on-chain and periodically updated. Thus miners are incentivized to collect as much relevant new data as quickly as possible.
Validation Process Details
In each validation round (every ~20 minutes):
At scale, this process creates a self-organizing data network: miners freely scrape data, validators continually audit it, and the reward mechanism ensures “quality rises to the top and value flows to contributors”.
Technical Architecture
OneOneOne is built as a Bittensor subnet using Node.js and Python. Key technical details:
Software Stack: The miner nodes and on-chain logic are implemented in Node.js (v18+) with associated npm packages. Python 3.12+ is used for AI/analysis components and environment management (via Conda). The GitHub repo shows a roughly 50/50 JS-Python codebase split.
Dependencies: Each node requires an API key for the web scraping service (Apify). Both miners and validators need a valid APIFY_TOKEN to run scraping agents against Google Maps, Yelp, etc.. A stake in TAO is also needed (per Bittensor rules) to run validator threads.
Data Sources: The system currently supports major review platforms. (The README explicitly mentions Google Maps and Yelp for location-based reviews, but it can be extended to other sites.)
Integration with Bittensor: As a Bittensor subnet, oneoneone inherits the network’s protocols. It uses Bittensor’s wallet/stake mechanism for metagraph participation. Rewards (TAO tokens) are distributed according to Bittensor’s mining/reward rules, augmented by the subnet’s buyback model. All communications between nodes occur via the Bittensor network (encrypted synapses), ensuring censorship resistance.
API Platform: The oneoneone.io website serves as the user interface and API endpoint. The data pipeline feeds into a live database that clients can query. (OneOneOne has built a subscription API layer on top of the metagraph.) No other proprietary protocols are used – oneoneone is largely built on standard open-source tools (Node, Python, Apify) within the Bittensor framework.
OneOneOne operates with three participant roles:
Miners: Run scraping agents (Node.js + Apify) that gather and format content from web platforms. Each miner has an API key (e.g. Apify token) and a stake in the subnet, and they launch data-collection jobs in response to challenges.
Validators: Issue and evaluate tasks. Every ~20 minutes a validator creates a synthetic challenge (e.g. “scrape all reviews for a random restaurant on Google Maps”). They randomly select up to 50 miners to participate, then verify the returned data. Validators check speed, format and accuracy (e.g. spot-checking responses against live platform data) to enforce quality. Validators also batch the outputs into the structured dataset.
API Consumers: Clients access the cleaned data via the oneoneone.io API. They pay subscription fees for this data.
The data pipeline proceeds in steps:
Data Scoring Mechanism
To rank miner submissions, oneoneone uses a cryptoeconomic scoring formula. Each miner’s score S is computed as:
S = 0.5 × V + 0.3 × T + 0.2 × R
where:
This 50/30/20 weighting (Volume/Speed/Recency) is enforced on-chain and periodically updated. Thus miners are incentivized to collect as much relevant new data as quickly as possible.
Validation Process Details
In each validation round (every ~20 minutes):
At scale, this process creates a self-organizing data network: miners freely scrape data, validators continually audit it, and the reward mechanism ensures “quality rises to the top and value flows to contributors”.
Technical Architecture
OneOneOne is built as a Bittensor subnet using Node.js and Python. Key technical details:
Software Stack: The miner nodes and on-chain logic are implemented in Node.js (v18+) with associated npm packages. Python 3.12+ is used for AI/analysis components and environment management (via Conda). The GitHub repo shows a roughly 50/50 JS-Python codebase split.
Dependencies: Each node requires an API key for the web scraping service (Apify). Both miners and validators need a valid APIFY_TOKEN to run scraping agents against Google Maps, Yelp, etc.. A stake in TAO is also needed (per Bittensor rules) to run validator threads.
Data Sources: The system currently supports major review platforms. (The README explicitly mentions Google Maps and Yelp for location-based reviews, but it can be extended to other sites.)
Integration with Bittensor: As a Bittensor subnet, oneoneone inherits the network’s protocols. It uses Bittensor’s wallet/stake mechanism for metagraph participation. Rewards (TAO tokens) are distributed according to Bittensor’s mining/reward rules, augmented by the subnet’s buyback model. All communications between nodes occur via the Bittensor network (encrypted synapses), ensuring censorship resistance.
API Platform: The oneoneone.io website serves as the user interface and API endpoint. The data pipeline feeds into a live database that clients can query. (OneOneOne has built a subscription API layer on top of the metagraph.) No other proprietary protocols are used – oneoneone is largely built on standard open-source tools (Node, Python, Apify) within the Bittensor framework.
No official team roster has been published. The GitHub organization oneoneone-io (which owns Subnet 111) shows no public members. However, public development activity indicates who’s involved: for example, GitHub commits have been authored by Bittensor community developers such as Roman (GitHub @basfroman-backup) – a San Francisco–based engineer – and others like ibraheem-abe, jarvis8x7b, etc. Many of these usernames correspond to known OpenTensor (Bittensor foundation) contributors. In short, the project appears to be led by a small core team within the Bittensor community, but individual names and affiliations have not been officially announced or published. (No LinkedIn/Twitter profiles for “oneoneone team” have been found.)
No official team roster has been published. The GitHub organization oneoneone-io (which owns Subnet 111) shows no public members. However, public development activity indicates who’s involved: for example, GitHub commits have been authored by Bittensor community developers such as Roman (GitHub @basfroman-backup) – a San Francisco–based engineer – and others like ibraheem-abe, jarvis8x7b, etc. Many of these usernames correspond to known OpenTensor (Bittensor foundation) contributors. In short, the project appears to be led by a small core team within the Bittensor community, but individual names and affiliations have not been officially announced or published. (No LinkedIn/Twitter profiles for “oneoneone team” have been found.)
Phase 1 (Kickoff): Implement 3 core job types, achieve ≥95% API success, and perform initial network benchmarking (current status: In Progress).
Phase 2 (AI Enhancement): Add LLM-powered features – automatic authenticity detection, intent and sentiment classification, and multi-language translation (status: Planned).
Phase 3 (Platform Launch): Release the public oneoneone.io dashboard, onboard early adopters, and optimize system performance (planned).
Phase 4 (Technical Expansion): Scale up to 10+ data “job” types, secure first paying customers, and deploy the on-chain buyback reward mechanism (planned).
Phase 5 (User Growth): Broaden the customer base, expand to 15+ job types, and roll out enhanced platform features (planned).
No dates are specified publicly; these milestones are indicated as ongoing or planned. As of mid-2025, Phase 1 work (core job types and reliability) is nearing completion, while later phases remain future goals. (If any recent updates or releases have occurred, they should be noted by checking the project’s GitHub or Bittensor community channels.)
Phase 1 (Kickoff): Implement 3 core job types, achieve ≥95% API success, and perform initial network benchmarking (current status: In Progress).
Phase 2 (AI Enhancement): Add LLM-powered features – automatic authenticity detection, intent and sentiment classification, and multi-language translation (status: Planned).
Phase 3 (Platform Launch): Release the public oneoneone.io dashboard, onboard early adopters, and optimize system performance (planned).
Phase 4 (Technical Expansion): Scale up to 10+ data “job” types, secure first paying customers, and deploy the on-chain buyback reward mechanism (planned).
Phase 5 (User Growth): Broaden the customer base, expand to 15+ job types, and roll out enhanced platform features (planned).
No dates are specified publicly; these milestones are indicated as ongoing or planned. As of mid-2025, Phase 1 work (core job types and reliability) is nearing completion, while later phases remain future goals. (If any recent updates or releases have occurred, they should be noted by checking the project’s GitHub or Bittensor community channels.)
Keep ahead of the Bittensor exponential development curve…
Subnet Alpha is an informational platform for Bittensor Subnets.
This site is not affiliated with the Opentensor Foundation or TaoStats.
The content provided on this website is for informational purposes only. We make no guarantees regarding the accuracy or currency of the information at any given time.
Subnet Alpha is created and maintained by The Realistic Trader. If you have any suggestions or encounter any issues, please contact us at [email protected].
Copyright 2024