Advanced knowledge graph tooling for AGI systems

SingularityNET April 16, 2025

Type

SingularityNET RFP

Total RFP Funding

$350,000 USD

Proposals

Awarded Projects

Completed & Awarded

The community and public are invited to view the full proposals and give feedback. During this time the RFP committee will doing their formal selection process to award winning proposals.

View Awarded Projects

Overview

Proposal Winners
Multiple
Max Funding / Proposal
$200,000 USD

Awarded Funding History

AWARDED PROJECTS	FUNDING AWARDED	ACTION
Autograph: Factoring Knowledge Graphs into Frames	$ 60000 USD	View
Augmenting BMKGs With Cheminformatics And CADD	$ 70000 USD	View
Dynamic KG Algorithms for Evolving AGI Systems	$ 87500 USD	View
Natural Language Explainability for Temporal KGs	$ 87500 USD	View
Hetzerk: AGI Reasoning and Decentralized Science	$ 50000 USD	View

Short summary

This RFP seeks the development of advanced tools and techniques for interfacing with, refining, and evaluating knowledge graphs that support reasoning in AGI systems. Projects may target any part of the graph lifecycle — from extraction to refinement to benchmarking — and should optionally support symbolic reasoning within the OpenCog Hyperon framework, including compatibility with the MeTTa language and MORK knowledge graph. Bids are expected to range from $10,000 - $200,000.

Main purpose

To develop advanced tools, algorithms, and/or frameworks for interfacing with, refining, augmentating, and evaluating knowledge graphs that support reasoning and symbolic processes within AGI systems. Solutions should aim to be domain-agnostic and, where possible, demonstrate integration with the MeTTa language and/or the MORK graph database for symbolic reasoning acceleration.

Long description

SingularityNET and its partners (OpenCog Foundation, TrueAGI) are building an ecosystem for Artificial General Intelligence (AGI) centered on the Hyperon framework. Within this architecture, knowledge graphs play a foundational role in enabling symbolic inference, analogical reasoning, causal understanding, and memory management. However, building and maintaining high-quality knowledge graphs remains a technically challenging and largely unsolved problem — particularly in dynamic or noisy domains like scientific research, open-ended learning, and evolving human knowledge. The goal of this RFP is to fund the development of open, extensible, and scalable tools to improve the full lifecycle of knowledge graphs in ways that stand to improve their integrity and utility in the context of neuro-symbolic AI. We welcome a wide range of proposals — including new extraction pipelines, refinement systems, dynamic update methods, schema-agnostic structuring tools, or benchmark frameworks to evaluate knowledge graph utility for AI reasoning. These systems could support one or more of the following use cases, with an emphasis on interface and utility over construction from scratch:

Creating and populating knowledge graphs from raw or semi-structured data (e.g., text, JSON, MeTTa expressions, publications).
Refining or distilling large, noisy, or redundant graphs into compact and semantically meaningful structures.
Detecting contradictions, inconsistencies, or anomalies in graph structure or contents (e.g., conflicting statements, invalid patterns).
Fact-checking or grounding LLM outputs using structured knowledge from knowledge graphs.
Designing or visualizing interfaces for how AGI systems reason with, act upon, or update knowledge graphs.
Annotating or weighting entities and relations (e.g., via confidence scores, semantic types, relationship types, etc.).
Enabling continuous graph updates with mechanisms for node/edge obsolescence detection.
Evaluating the effectiveness of knowledge graphs in supporting AGI reasoning tasks such as multi-hop question answering, analogical retrieval, and hypothesis generation.
Exporting or converting knowledge graphs into formats consumable by other symbolic tools like MeTTa or the MORK system.

Solutions that can demonstrate compatibility or integration with the MORK system — a fast hypergraph backend for MeTTa — are encouraged. MORK’s ability to process billion-scale symbolic spaces at interactive speeds opens up new possibilities for real-time reasoning, retrieval, and cognitive synergy within Hyperon-based agents. We encourage proposals that are modular and reusable across domains, and that can demonstrate either standalone value or close integration with the existing SingularityNET ecosystem. Potential contributors are also welcome to collaborate or coordinate with the developers of MeTTa, Atomspace, or MORK to ensure alignment and future extensibility. Background & experience: A significant part of reviewing proposals goes into evaluating the ability of a team to execute the work. Please provide in as much detail as possible related experience and accolades, and supply links to anything we can read such as published work, github, etc.

Functional Requirements

Must have

A clear technical plan for developing or enhancing tools that improve the quality of knowledge graphs and the way we work with them in the context of neuro-symbolic AI.
Outputs that can be tested or demonstrated on real datasets (preferably in scientific, technical, or cross-domain contexts).
Documentation or outputs sufficient to be adopted or reused by third-party developers.

Should have

Support for integration into symbolic systems like MeTTa or Atomspace.
Benchmarks or evaluations showing performance, compactness, noise tolerance, and reasoning support.

Could have

Support for graph integrity tools (e.g., anomaly detection, contradiction resolution, edge validation).
Tools for grounding or fact-checking LLM outputs using graph lookups.
Exploratory interfaces or APIs that demonstrate how AGI systems might interact with, navigate, or co-evolve with KGs.
Tools to mitigate bias and evaluate the confidence of data and data sources.
Demonstration or testing using MORK as a graph backend.
Tools or schemas for structuring LLM-extracted knowledge into usable graph formats.
User interfaces or APIs to allow LLMs or external tools to query or modify the graph.
Support for live or streaming updates to knowledge graphs over time.

Won’t have

Focus on pure visualization without functionality.
Use of proprietary or closed-source formats that cannot be ported or reused.

Non-functional Requirements

Must Have

Tools or components that improve knowledge graph quality (e.g., structure, accuracy, interpretability, or query-ability) in the context of their utility for neuro-symbolic AI systems.
Modular and reusable code that can be easily adopted by other developers or integrated into broader systems.
Open data formats (e.g., JSON, MeTTa S-expressions, or common graph formats).
Reasonable performance for prototyping (e.g., works on medium-sized graphs with hundreds of thousands of nodes/edges on standard hardware).
Developer documentation and example usage.

Should Have

Support for graph enrichment features like typed edges, edge weights, timestamps, or confidence scores.
Evaluation or benchmarking tools to assess graph quality, utility, or reasoning effectiveness.
Partial or planned support for larger-scale graphs (e.g., >10M nodes/edges).
APIs or scripts for batch processing and integration with other tools (e.g., extract-transform-load workflows).

Could Have

Integration with the MeTTa language or compatibility with MORK (e.g., exporting/importing graph data, or using MORK as a backend).
Interactive or semantic query tools for exploring or testing knowledge graphs.
Streaming or dynamic update support (add/remove/refine nodes/edges on the fly).
Tools for formatting graphs for LLM consumption or hybrid symbolic-LLM workflows.
Containerized deployment, REST interface, or CLI for easy testing and reuse.

Team requirements

Teams should include at least one developer or researcher with prior experience in graph algorithms, knowledge representation, or NLP.
Familiarity with MeTTa, MORK, or symbolic AI frameworks is a plus but not required.
Comfortable working in an open-source context (e.g., pushing to GitHub, using open licenses).
Clear communication and progress reporting during milestones is expected (e.g., through GitHub issues, milestone docs, or biweekly check-ins if requested).
Prior examples of relevant work (code repos, papers, benchmarks, etc.) are strongly encouraged and will be considered in evaluation.

Main evaluation criteria

Alignment with requirements and objective

Does the proposal meet the requirements and advances the objectives of the RFP.

Pre-existing R&D

Has the team previously done similar or related research or development work in other platforms / languages / contexts?

Team competence

Does the team have relevant skills?

Cost

Does the proposal offer good value for money?

Timeline

Does the proposal include a set of clearly defined milestones?
We highly recommend submitting proposals with project milestones along the lines of the following:
- Milestone 1: Submit a thorough research plan outlining and detailing the approach and work to be done. Deliverables: detailed research plan, agile breakdown of tasks with timeline, and framework design. 20% of grant
- Milestone 2: Complete initial development of the framework definitively showing implementation of the conceptual underpinnings of the RFP along with preliminary testing. Deliverables: draft implementation, initial testing results, and analysis against standard benchmarks. 40% of grant
- Milestone 3: Submit all final materials as committed to in the grant proposal.Deliverables (as applicable): final report with performance analysis, code, framework demonstration, documentation, recommendations, websites, etc. 40% of grant

Other resources

SingularityNET technology links
MeTTa language documentation
MORK github
SingularityNET holds MeTTa study group calls every other week. Proposers are welcome to attend for support from our researchers and community.
Recurring Hyperon study group calls for community are currently being planned. These will cover MOSES, ECAN, PLN, and other key components of the OpenCog and PRIMUS Hyperon cognitive architectures.
Access to the SingularityNET World Mattermost server, with a dedicated channel for discussion and support among the RFP-winning teams and SingularityNET resources.