There are very established graphical programming languages in the art-scene. They are easy to get in to, but become messy for complex projects.
For the erst stuff think they are a bit like jupyter notebooks, but even better: everything compiles in real time and you can see it working. On the other hand it quickly just becomes a spaghetti mess..
For everything else you need things with a limited scope. A single state to track (workflows, conversations) or really basic logic and strong abstractions (mangling two APIs together). Beyond that you need a programmer anyways and his gain from a GUI is limited.
Generally there are a few things hard to represent, e.g. abstracting and recycling code (writing functions), parallel processes, state and highly interconnected things.
A few examples for the curious:
For Visuals:
- quartz composer (old by now) http://www.mactricksandtips.com/wp-content/uploads/2008/03/n...
- touch designer (modern, very nice nesting, you can zoom in to groups): https://youtu.be/hbZjgHSCAPI?t=49
Music:
Pure Data and maxMSP (not strictly for just for music): https://youtu.be/rTQgfhsQ7xo
Bitwig Grid (very skeumorphic, yet one of the modern "modular" ones. I dig the look of the droopy bezier curves though): https://youtu.be/dNdhbHGeHPw
More recent and really interesting to me is "no-code" environments that are now gaining traction.
Business logic:
BPMN + Camunda (you still need to code everything in text, but you can shuffle the flow around afterwards): https://youtu.be/HxtZf5VD6lQ?t=625
No-code API plugging:
Appmixer: https://uploads-ssl.webflow.com/5a9d00dba5e9fa00010cb403/5c8...
AI-assisted chatbots:
Cognigy: https://youtu.be/QSJ-nTwjn-c?t=1525"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Why isn\u2019t visual programming a bigger thing?"}},"_tags":["comment","author_anotheryou","story_23252448"],"author":"anotheryou","comment_text":"2D space is quite limited in arrangement I think (and with 3D you always have occlusion).
There are very established graphical programming languages in the art-scene. They are easy to get in to, but become messy for complex projects.
For the erst stuff think they are a bit like jupyter notebooks, but even better: everything compiles in real time and you can see it working. On the other hand it quickly just becomes a spaghetti mess..
For everything else you need things with a limited scope. A single state to track (workflows, conversations) or really basic logic and strong abstractions (mangling two APIs together). Beyond that you need a programmer anyways and his gain from a GUI is limited.
Generally there are a few things hard to represent, e.g. abstracting and recycling code (writing functions), parallel processes, state and highly interconnected things.
A few examples for the curious:
For Visuals:
- quartz composer (old by now) http://www.mactricksandtips.com/wp-content/uploads/2008/03/n...
- touch designer (modern, very nice nesting, you can zoom in to groups): https://youtu.be/hbZjgHSCAPI?t=49
Music:
Pure Data and maxMSP (not strictly for just for music): https://youtu.be/rTQgfhsQ7xo
Bitwig Grid (very skeumorphic, yet one of the modern "modular" ones. I dig the look of the droopy bezier curves though): https://youtu.be/dNdhbHGeHPw
More recent and really interesting to me is "no-code" environments that are now gaining traction.
Business logic:
BPMN + Camunda (you still need to code everything in text, but you can shuffle the flow around afterwards): https://youtu.be/HxtZf5VD6lQ?t=625
No-code API plugging:
Appmixer: https://uploads-ssl.webflow.com/5a9d00dba5e9fa00010cb403/5c8...
AI-assisted chatbots:
Cognigy: https://youtu.be/QSJ-nTwjn-c?t=1525","created_at":"2020-05-21T11:27:08Z","created_at_i":1590060428,"objectID":"23257129","parent_id":23252448,"story_id":23252448,"story_title":"Ask HN: Why isn\u2019t visual programming a bigger thing?","updated_at":"2024-09-20T06:15:12Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"standup75"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Launching Beta version of a free alternative to Coinigy"},"url":{"matchLevel":"none","matchedWords":[],"value":"http://silexapp.com"}},"_tags":["story","author_standup75","story_17316137"],"author":"standup75","children":[17316140,17316265,17316276],"created_at":"2018-06-14T23:00:51Z","created_at_i":1529017251,"num_comments":6,"objectID":"17316137","points":3,"story_id":17316137,"title":"Launching Beta version of a free alternative to Coinigy","updated_at":"2024-09-20T02:33:10Z","url":"http://silexapp.com"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"brentdev"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Show HN: Coinigy \u2013 Professional Tools for Cryptocurrency Traders"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"https://www.coinigy.com/"}},"_tags":["story","author_brentdev","story_8742440","show_hn"],"author":"brentdev","created_at":"2014-12-12T20:15:45Z","created_at_i":1418415345,"num_comments":0,"objectID":"8742440","points":2,"story_id":8742440,"story_text":"","title":"Show HN: Coinigy \u2013 Professional Tools for Cryptocurrency Traders","updated_at":"2024-09-19T21:25:28Z","url":"https://www.coinigy.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"madara8865"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"CogniGraph \u2013 describe a scenario, LLM picks a brain lobe and neuromodulator"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"https://github.com/madara88645/Cognigraph"}},"_tags":["story","author_madara8865","story_48396828"],"author":"madara8865","children":[48396829],"created_at":"2026-06-04T10:52:46Z","created_at_i":1780570366,"num_comments":0,"objectID":"48396828","points":2,"story_id":48396828,"title":"CogniGraph \u2013 describe a scenario, LLM picks a brain lobe and neuromodulator","updated_at":"2026-06-05T14:16:48Z","url":"https://github.com/madara88645/Cognigraph"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rexreed"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Cognilytica Tested Voice Assistants and You Might Be Surprised by Results(podcast)"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"http://www.cognilytica.com/2018/02/28/ai-today-podcast-26-cognilytica-tested-voice-assistants-might-surprised-results/"}},"_tags":["story","author_rexreed","story_16487008"],"author":"rexreed","created_at":"2018-02-28T21:14:39Z","created_at_i":1519852479,"num_comments":0,"objectID":"16487008","points":2,"story_id":16487008,"title":"Cognilytica Tested Voice Assistants and You Might Be Surprised by Results(podcast)","updated_at":"2024-09-20T02:08:11Z","url":"http://www.cognilytica.com/2018/02/28/ai-today-podcast-26-cognilytica-tested-voice-assistants-might-surprised-results/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"soco"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Cognify: Revolutionary Prison Concept Uses AI and Brain Implants To"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"https://www.sciencetimes.com/articles/50929/20240626/cognify-revolutionary-prison-concept-uses-ai-brain-implants-fast-track.htm"}},"_tags":["story","author_soco","story_40854499"],"author":"soco","children":[40854509,40854612,40854635],"created_at":"2024-07-02T08:02:17Z","created_at_i":1719907337,"num_comments":4,"objectID":"40854499","points":1,"story_id":40854499,"title":"Cognify: Revolutionary Prison Concept Uses AI and Brain Implants To","updated_at":"2024-09-20T17:23:50Z","url":"https://www.sciencetimes.com/articles/50929/20240626/cognify-revolutionary-prison-concept-uses-ai-brain-implants-fast-track.htm"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hisabness"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Looking for an alternative to Coinigy that is not buggy, and enables tracking of trades not conducted via their API. Any thoughts?"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Ask HN: Coinigy Alternative for Crypto Administration and Accounting?"}},"_tags":["story","author_hisabness","story_15826293","ask_hn"],"author":"hisabness","children":[15827922],"created_at":"2017-12-01T19:00:11Z","created_at_i":1512154811,"num_comments":2,"objectID":"15826293","points":1,"story_id":15826293,"story_text":"Looking for an alternative to Coinigy that is not buggy, and enables tracking of trades not conducted via their API. Any thoughts?","title":"Ask HN: Coinigy Alternative for Crypto Administration and Accounting?","updated_at":"2024-09-20T01:40:09Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"reynabhyankar"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Cognify: Automated Quality+Cost Optimization for GenAI Workflow"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"https://github.com/GenseeAI/cognify"}},"_tags":["story","author_reynabhyankar","story_42248836"],"author":"reynabhyankar","children":[42248837],"created_at":"2024-11-26T19:07:09Z","created_at_i":1732648029,"num_comments":1,"objectID":"42248836","points":1,"story_id":42248836,"title":"Cognify: Automated Quality+Cost Optimization for GenAI Workflow","updated_at":"2024-11-26T19:10:05Z","url":"https://github.com/GenseeAI/cognify"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vasa_"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Hey HN! We're Vasilije, Laszlo and Lazar, the authors of a new paper and part of https://www.cognee.ai. cognee let\u2019s you build memory layers for AI applications and agents, allowing them to personalize results, connect various data sources, and add custom rules. This enables AI apps to deliver increasingly accurate responses, we reached almost 90% on standard industry benchmarks as you can see here https://github.com/topoteretes/cognee/tree/main/evals and our paper can be accessed at: https://arxiv.org/abs/2505.24478 and collab is in the repository
LLMs don't remember context well\u2014they can\u2019t ingest your data and keep it in memory. This limitation leads to lacking interactions, a lack of accuracy, and the inability to connect your data sources cheaply because developers must include long, unmanaged context in every prompt.
When we were building RAGs we saw that we can\u2019t find the data we need and that there are too many knobs to turn in RAG frameworks. We had to tweak many parameters and also could not specify the rules we wanted the data to follow. No ontologies, no rulesets, no state or good data engineering practices and a lot of manual work. That\u2019s why cognee
cognee builds memory that combines graph, vector, and relational stores. Here is how it works:
Adding data: When you use cognee with your AI App, it can take in any message, string, S3 bucket or even a relational database and automatically ingest it
Managing information: cognee sorts this information into semantic graph: - It extracts entities and connections between things. - It embeds the data in the vector store, It enriches data with custom ontologies you provide, that help ground the graphs and make them more reliable. - The overall information is stored in many layers of a graph and vector store that allows for finding similar information later using a variety of types of searches.
Retrieving data: When given an input query, cognee searches for and retrieves related stored information by leveraging a combination of graph traversal techniques, vector similarity, and COT techniques. It can use the internal benchmarking system to make sure that your pipelines are returning only accurate data when you need it.
---
cognee introduces a self-improving group of memory layers that cover various topics, data sources and can be customized. This reduces the need to build everything from scratch, and you can use our primitives to get started and move more quickly. On the other hand, you can do everything yourself, from start to end. We\u2019ve designed the system to be modular and extensible.
We\u2019ve open-sourced cognee \u2014specifically the framework and various vector and graph database adapters, as well as our default data pipeline, cognify\u2014under the Apache 2.0 license. This includes the ability to add, cognify, and retrieve data within your AI applications and also extend it with custom components that are just pure Python.
However, many keep features that are optimized for production use, as a part of our paid platform. We release these functionalities too, including in the last few weeks permission management + distributed pipelines(dev). These are a part of our open-source package and are available to those who in production environments need things like rate limiting and credential management. We will release a paid offering for developers to get easy API access to our platform.
Our automation tooling allows us to optimize the pipelines and find the best combination of parameters that answer questions our stakeholders have.
We\u2019d love to hear what you think! Please feel free to try our demo, check out the code, read the research paper, and share thoughts or suggestions with us. Your feedback will help shape where we take cognee from here!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Cognee \u2013 Open-Source AI Memory Layer That Remembers Context"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/topoteretes/cognee"}},"_tags":["story","author_vasa_","story_44169594","show_hn"],"author":"vasa_","children":[44171866],"created_at":"2025-06-03T13:05:15Z","created_at_i":1748955915,"num_comments":2,"objectID":"44169594","points":9,"story_id":44169594,"story_text":"Hey HN! We're Vasilije, Laszlo and Lazar, the authors of a new paper and part of https://www.cognee.ai. cognee let\u2019s you build memory layers for AI applications and agents, allowing them to personalize results, connect various data sources, and add custom rules. This enables AI apps to deliver increasingly accurate responses, we reached almost 90% on standard industry benchmarks as you can see here https://github.com/topoteretes/cognee/tree/main/evals and our paper can be accessed at: https://arxiv.org/abs/2505.24478 and collab is in the repository
LLMs don't remember context well\u2014they can\u2019t ingest your data and keep it in memory. This limitation leads to lacking interactions, a lack of accuracy, and the inability to connect your data sources cheaply because developers must include long, unmanaged context in every prompt.
When we were building RAGs we saw that we can\u2019t find the data we need and that there are too many knobs to turn in RAG frameworks. We had to tweak many parameters and also could not specify the rules we wanted the data to follow. No ontologies, no rulesets, no state or good data engineering practices and a lot of manual work. That\u2019s why cognee
cognee builds memory that combines graph, vector, and relational stores. Here is how it works:
Adding data: When you use cognee with your AI App, it can take in any message, string, S3 bucket or even a relational database and automatically ingest it
Managing information: cognee sorts this information into semantic graph: - It extracts entities and connections between things. - It embeds the data in the vector store, It enriches data with custom ontologies you provide, that help ground the graphs and make them more reliable. - The overall information is stored in many layers of a graph and vector store that allows for finding similar information later using a variety of types of searches.
Retrieving data: When given an input query, cognee searches for and retrieves related stored information by leveraging a combination of graph traversal techniques, vector similarity, and COT techniques. It can use the internal benchmarking system to make sure that your pipelines are returning only accurate data when you need it.
---
cognee introduces a self-improving group of memory layers that cover various topics, data sources and can be customized. This reduces the need to build everything from scratch, and you can use our primitives to get started and move more quickly. On the other hand, you can do everything yourself, from start to end. We\u2019ve designed the system to be modular and extensible.
We\u2019ve open-sourced cognee \u2014specifically the framework and various vector and graph database adapters, as well as our default data pipeline, cognify\u2014under the Apache 2.0 license. This includes the ability to add, cognify, and retrieve data within your AI applications and also extend it with custom components that are just pure Python.
However, many keep features that are optimized for production use, as a part of our paid platform. We release these functionalities too, including in the last few weeks permission management + distributed pipelines(dev). These are a part of our open-source package and are available to those who in production environments need things like rate limiting and credential management. We will release a paid offering for developers to get easy API access to our platform.
Our automation tooling allows us to optimize the pipelines and find the best combination of parameters that answer questions our stakeholders have.
We\u2019d love to hear what you think! Please feel free to try our demo, check out the code, read the research paper, and share thoughts or suggestions with us. Your feedback will help shape where we take cognee from here!","title":"Show HN: Cognee \u2013 Open-Source AI Memory Layer That Remembers Context","updated_at":"2026-06-26T19:29:14Z","url":"https://github.com/topoteretes/cognee"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vasa_"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Hi,\nWe are building an open-source framework for loading and structuring LLM context to create accurate and explainable LLM answers using knowledge graphs and vector stores.
We built the tool with four main concepts in mind:
1. Loader -> uses dlt in the backend to load and structure the data
2. Cognify step -> creates a graph with summaries, labels and factoids that are interconnected across the documents and stored as a representation in the vector store
3. Optimizer -> Uses DSPy to optimize LLM queries, and we plan to extend it to most of the knobs we can turn, like chunking etc.
4. Search -> allows for searching using search types supported in graph stores (ex. Neo4j) or hybrid, BM25, or other search types available in vector stores.
We are quite early with the product but we would love to hear feedback on what we can improve."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Auto-optimizing deterministic LLM outputs using knowledge graphs"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/topoteretes/cognee"}},"_tags":["story","author_vasa_","story_40117217","show_hn"],"author":"vasa_","created_at":"2024-04-22T18:25:59Z","created_at_i":1713810359,"num_comments":0,"objectID":"40117217","points":7,"story_id":40117217,"story_text":"Hi,\nWe are building an open-source framework for loading and structuring LLM context to create accurate and explainable LLM answers using knowledge graphs and vector stores.
We built the tool with four main concepts in mind:
1. Loader -> uses dlt in the backend to load and structure the data
2. Cognify step -> creates a graph with summaries, labels and factoids that are interconnected across the documents and stored as a representation in the vector store
3. Optimizer -> Uses DSPy to optimize LLM queries, and we plan to extend it to most of the knobs we can turn, like chunking etc.
4. Search -> allows for searching using search types supported in graph stores (ex. Neo4j) or hybrid, BM25, or other search types available in vector stores.
We are quite early with the product but we would love to hear feedback on what we can improve.","title":"Show HN: Auto-optimizing deterministic LLM outputs using knowledge graphs","updated_at":"2024-09-20T16:50:10Z","url":"https://github.com/topoteretes/cognee"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vasa_"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Hey there HN! We\u2019re Vasilije, Boris, and Laszlo, and we\u2019re excited to introduce cognee, an open-source Python library that approaches building evolving semantic memory using knowledge graphs + data pipelines
Before we built cognee, Vasilije(B Economics and Clinical Psychology) worked at a few unicorns (Omio, Zalando, Taxfix), while Boris managed large-scale applications in production at Pera and StuDocu. Laszlo joined after getting his PhD in Graph Theory at the University of Szeged.
Using LLMs to connect to large datasets (RAG) has been popularized and has shown great promise. Unfortunately, this approach doesn\u2019t live up to the hype.
Let\u2019s assume we want to load a large repository from GitHub to a vector store.\nConnectingfiles in larger systems with RAG would fail because a fixed RAG limit is too constraining in longer dependency chains. While we need results that are aware of the context of the whole repository, RAG\u2019s similarity-based retrieval does not capture the full context of interdependent files spread across the repository.
This approach allows cognee to retrieve all relevant and correct context at inference time. For example, if `function A` in one file calls `function B` in another file, which calls `function C` in a third file, all code and summaries that further explain their position and purpose in that chain are served as context. As a result, the system has complete visibility into how different code parts work together within the repo.
Last year, Microsoft took a leap published GraphRAG - i.e. RAG with Knowledge Graphs. We think it is the right direction.\nOur initial ideas were similar to this paper and this got some attention on Twitter (https://x.com/tricalt/status/1722216426709365024)
Over time we understood we needed tooling to create dynamically evolving groups of graphs, cross-connected and evaluated together.\nOur tool is named after a process called cognification. We prefer the definition that Vakalo (1978) uses to explain that cognify represents "building a fitting (mental) picture"
We believe that agents of tomorrow will require a correct dynamic \u201cmental picture\u201d or context to operate in a rapidly evolving landscape.
To address this, we built ECL pipelines, where we do the following:\n- Extract data from various sources using dlt and existing frameworks\n- Cognify - create a graph/vector representation of the data\n- Load - store the data in the vector (in this case our partner FalkorDB), graph, and relational stores
We can also continuously feed the graph with new information, and when testing this approach we found that on HotpotQA, with human labeling, we achieved 87% answer accuracy (https://docs.cognee.ai/evaluations).
To show how the approach works we did an integration with continue.dev and built a codegraph
Here is how codegraph was implemented: \nWe're explicitly including repository structure details and integrating custom dependency graph versions. Think of it as a more insightful way to understand your codebase's architecture.\nBy transforming dependency graphs into knowledge graphs, we're creating a quick, graph-based version of tools like tree-sitter. This means faster and more accurate code analysis.\nWe worked on modeling causal relationships within code and enriching them with LLMs. This helps you understand how different parts of your code influence each other.\nWe created graph skeletons in memory which allows us to perform various operations on graphs and power custom retrievers.
If you want to integrate cognee into your systems or have a look at codegraph, our GitHub repository is (https://github.com/topoteretes/cognee)
Thank you for reading! We\u2019re definitely early and welcome your ideas and experiences as it relates to agents, graphs, evals, and human+LLM memory."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Cognee \u2013 Turn RAG and GraphRAG into custom dynamic semantic memory"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/topoteretes/cognee"}},"_tags":["story","author_vasa_","story_43031915","show_hn"],"author":"vasa_","children":[43034688],"created_at":"2025-02-13T01:59:26Z","created_at_i":1739411966,"num_comments":1,"objectID":"43031915","points":6,"story_id":43031915,"story_text":"Hey there HN! We\u2019re Vasilije, Boris, and Laszlo, and we\u2019re excited to introduce cognee, an open-source Python library that approaches building evolving semantic memory using knowledge graphs + data pipelines
Before we built cognee, Vasilije(B Economics and Clinical Psychology) worked at a few unicorns (Omio, Zalando, Taxfix), while Boris managed large-scale applications in production at Pera and StuDocu. Laszlo joined after getting his PhD in Graph Theory at the University of Szeged.
Using LLMs to connect to large datasets (RAG) has been popularized and has shown great promise. Unfortunately, this approach doesn\u2019t live up to the hype.
Let\u2019s assume we want to load a large repository from GitHub to a vector store.\nConnectingfiles in larger systems with RAG would fail because a fixed RAG limit is too constraining in longer dependency chains. While we need results that are aware of the context of the whole repository, RAG\u2019s similarity-based retrieval does not capture the full context of interdependent files spread across the repository.
This approach allows cognee to retrieve all relevant and correct context at inference time. For example, if `function A` in one file calls `function B` in another file, which calls `function C` in a third file, all code and summaries that further explain their position and purpose in that chain are served as context. As a result, the system has complete visibility into how different code parts work together within the repo.
Last year, Microsoft took a leap published GraphRAG - i.e. RAG with Knowledge Graphs. We think it is the right direction.\nOur initial ideas were similar to this paper and this got some attention on Twitter (https://x.com/tricalt/status/1722216426709365024)
Over time we understood we needed tooling to create dynamically evolving groups of graphs, cross-connected and evaluated together.\nOur tool is named after a process called cognification. We prefer the definition that Vakalo (1978) uses to explain that cognify represents "building a fitting (mental) picture"
We believe that agents of tomorrow will require a correct dynamic \u201cmental picture\u201d or context to operate in a rapidly evolving landscape.
To address this, we built ECL pipelines, where we do the following:\n- Extract data from various sources using dlt and existing frameworks\n- Cognify - create a graph/vector representation of the data\n- Load - store the data in the vector (in this case our partner FalkorDB), graph, and relational stores
We can also continuously feed the graph with new information, and when testing this approach we found that on HotpotQA, with human labeling, we achieved 87% answer accuracy (https://docs.cognee.ai/evaluations).
To show how the approach works we did an integration with continue.dev and built a codegraph
Here is how codegraph was implemented: \nWe're explicitly including repository structure details and integrating custom dependency graph versions. Think of it as a more insightful way to understand your codebase's architecture.\nBy transforming dependency graphs into knowledge graphs, we're creating a quick, graph-based version of tools like tree-sitter. This means faster and more accurate code analysis.\nWe worked on modeling causal relationships within code and enriching them with LLMs. This helps you understand how different parts of your code influence each other.\nWe created graph skeletons in memory which allows us to perform various operations on graphs and power custom retrievers.
If you want to integrate cognee into your systems or have a look at codegraph, our GitHub repository is (https://github.com/topoteretes/cognee)
Thank you for reading! We\u2019re definitely early and welcome your ideas and experiences as it relates to agents, graphs, evals, and human+LLM memory.","title":"Show HN: Cognee \u2013 Turn RAG and GraphRAG into custom dynamic semantic memory","updated_at":"2026-06-26T19:29:14Z","url":"https://github.com/topoteretes/cognee"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"kennethc"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Hi HN! I'm one of the creators of MoreLogin and we're happy (and once again very nervous) to release this on HN. MoreLogin is an anti-detection browser, and the fact that there are already many anti-detection browsers on the market, even though we've been iterating on our anti-association capabilities to reach the top tier of the market, doesn't seem to feel like much to users. We have also thought about developing features specifically for crypto users, such as specifically developing trading platforms and wallets that can connect to mt5 and coinigy, and making automated trading robots to support them, as well as developing features specifically for marketers, such as ensuring the success rate of account registration. We think it's a cool idea, but it's all just based on our own ideas and we really don't know if it's useful and if people will pay for it. So we wanted to see what needs people had for this kind of tool, this kind of browser, and what we could go to meet them.
Here's the intro video we made a while back - https://www.youtube.com/watch?v=hW4O8yXe4r4
Any feedback or questions are welcome! :) Thanks!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: What do you want or expect from an antidetect browser"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.morelogin.com"}},"_tags":["story","author_kennethc","story_34605521","show_hn"],"author":"kennethc","created_at":"2023-02-01T02:22:33Z","created_at_i":1675218153,"num_comments":0,"objectID":"34605521","points":1,"story_id":34605521,"story_text":"Hi HN! I'm one of the creators of MoreLogin and we're happy (and once again very nervous) to release this on HN. MoreLogin is an anti-detection browser, and the fact that there are already many anti-detection browsers on the market, even though we've been iterating on our anti-association capabilities to reach the top tier of the market, doesn't seem to feel like much to users. We have also thought about developing features specifically for crypto users, such as specifically developing trading platforms and wallets that can connect to mt5 and coinigy, and making automated trading robots to support them, as well as developing features specifically for marketers, such as ensuring the success rate of account registration. We think it's a cool idea, but it's all just based on our own ideas and we really don't know if it's useful and if people will pay for it. So we wanted to see what needs people had for this kind of tool, this kind of browser, and what we could go to meet them.
Here's the intro video we made a while back - https://www.youtube.com/watch?v=hW4O8yXe4r4
Any feedback or questions are welcome! :) Thanks!","title":"Show HN: What do you want or expect from an antidetect browser","updated_at":"2024-09-20T13:10:16Z","url":"https://www.morelogin.com"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dvanduzer"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"What's M5?
I had a short gig at Cognilytics in 2010, and I can't figure out what any version of it has to do with Kawasaki Robotics, but..."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"A Room Where Executives Go to Get Help from IBM\u2019s Watson"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"http://www.technologyreview.com/news/529606/a-room-where-executives-go-to-get-help-from-ibms-watson"}},"_tags":["comment","author_dvanduzer","story_8154071"],"author":"dvanduzer","comment_text":"What's M5?
I had a short gig at Cognilytics in 2010, and I can't figure out what any version of it has to do with Kawasaki Robotics, but...","created_at":"2014-08-08T21:32:39Z","created_at_i":1407533559,"objectID":"8155158","parent_id":8154588,"points":null,"story_id":8154071,"story_title":"A Room Where Executives Go to Get Help from IBM\u2019s Watson","story_url":"http://www.technologyreview.com/news/529606/a-room-where-executives-go-to-get-help-from-ibms-watson","updated_at":"2023-09-06T22:09:08Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"MalcolmPF"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":">\u201cI recommend eliminating Kawasaki Robotics.\u201d When Watson was asked to explain, it simply added. \u201cIt is inferior to Cognilytics in every way.\u201d
Getting some serious M5 vibes right now."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"A Room Where Executives Go to Get Help from IBM\u2019s Watson"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"http://www.technologyreview.com/news/529606/a-room-where-executives-go-to-get-help-from-ibms-watson"}},"_tags":["comment","author_MalcolmPF","story_8154071"],"author":"MalcolmPF","children":[8155158,8155843],"comment_text":">\u201cI recommend eliminating Kawasaki Robotics.\u201d When Watson was asked to explain, it simply added. \u201cIt is inferior to Cognilytics in every way.\u201d
Getting some serious M5 vibes right now.","created_at":"2014-08-08T19:23:21Z","created_at_i":1407525801,"objectID":"8154588","parent_id":8154071,"points":null,"story_id":8154071,"story_title":"A Room Where Executives Go to Get Help from IBM\u2019s Watson","story_url":"http://www.technologyreview.com/news/529606/a-room-where-executives-go-to-get-help-from-ibms-watson","updated_at":"2023-09-06T22:09:03Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"reynabhyankar"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Hi everyone! I'm Reyna, a PhD student working on systems for machine learning.
I want to share an exciting open-source project my team has built: Cognify. Cognify is a multi-faceted optimization tool that automatically enhances generation quality and reduces execution costs for generative AI workflows written in LangChain, DSPy, and Python. Cognify helps you evaluate and refine your workflows at any stage of development. Use it to test and enhance workflows you\u2019ve finished building or to analyze your current workflow\u2019s potential.
Key highlights:\n- Workflow generation quality improvement by up to 48%\n- Workflow execution cost reduction by up to 9x\n- Multiple optimized workflow versions with quality-cost combinations for you to choose\n- Automatic model selection, prompt enhancing, and workflow structure optimization
Get Cognify at https://github.com/GenseeAI/cognify and read more at https://mlsys.wuklab.io/posts/cognify/. Would love to hear your feedback and get your contributions!"},"story_title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"Cognify: Automated Quality+Cost Optimization for GenAI Workflow"},"story_url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"https://github.com/GenseeAI/cognify"}},"_tags":["comment","author_reynabhyankar","story_42248836"],"author":"reynabhyankar","comment_text":"Hi everyone! I'm Reyna, a PhD student working on systems for machine learning.
I want to share an exciting open-source project my team has built: Cognify. Cognify is a multi-faceted optimization tool that automatically enhances generation quality and reduces execution costs for generative AI workflows written in LangChain, DSPy, and Python. Cognify helps you evaluate and refine your workflows at any stage of development. Use it to test and enhance workflows you\u2019ve finished building or to analyze your current workflow\u2019s potential.
Key highlights:\n- Workflow generation quality improvement by up to 48%\n- Workflow execution cost reduction by up to 9x\n- Multiple optimized workflow versions with quality-cost combinations for you to choose\n- Automatic model selection, prompt enhancing, and workflow structure optimization
Get Cognify at https://github.com/GenseeAI/cognify and read more at https://mlsys.wuklab.io/posts/cognify/. Would love to hear your feedback and get your contributions!","created_at":"2024-11-26T19:07:09Z","created_at_i":1732648029,"objectID":"42248837","parent_id":42248836,"story_id":42248836,"story_title":"Cognify: Automated Quality+Cost Optimization for GenAI Workflow","story_url":"https://github.com/GenseeAI/cognify","updated_at":"2024-11-26T19:10:05Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"karaterobot"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["cognigy"],"value":"> In private conversations, Mr. Altman has compared the world\u2019s data centers to electricity, according to three people close to the discussions. As the availability of electricity became more widespread, people found better ways of using it. Mr. Altman hoped to do the same with data centers and eventually make A.I. technologies flow like electricity.
Reminds me of a quote I saved from Kevin Kelly's book The Inevitable, which was published in 2016. I saved it because it sounded so absurd at the time. To be clear, it sort of still does sound absurd to me, who remain moderately skeptical about AI (though admittedly less so than in 2016, largely because of the following sentence). What's very serious is how much money and old-fashioned brainpower is going into making this future real. I don't know whether the metaphor was arrived at independently, derives from the same source, or is just an obvious way of thinking about the world when you've drunk the right flavor of Kool-Aid.
> Amid all this activity, of a picture of our AI future is coming into view, and it is not the HAL 9000\u2014a discrete machine animated by a charismatic (yet potentially homicidal) humanlike consciousness\u2014or a Singularitan rapture of superintelligence. The AI on the horizon looks more like Amazon Web Services\u2014cheap, reliable, industrial-grade digital smartness running behind everything, and almost invisible except when it blinks off. This common utility will serve you as much IQ as you want but no more than you need. You\u2019ll simply plug into the grid and get AI as if it was electricity. It will enliven inert objects, much as electricity did more than a century past. Three generations ago, many a tinkerer struck it rich by taking a tool and making an electric version. Take a manual pump; electrify it. Find a hand-wringer washer; electrify it. The entrepreneurs didn\u2019t need to generate the electricity; they brought it from the grid and used it to automate the previously manual. Now everything that we formerly electrified we will cognify. There is almost nothing we can think of that cannot be made new, different, or more valuable by infusing it with some extra IQ. In fact, the business plans of the next 10,000 startups are easy to forecast: Take X and add AI. Find something that can be made better by adding online smartness to it."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Behind OpenAI's plan to make A.I. flow like electricity"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.nytimes.com/2024/09/25/business/openai-plan-electricity.html"}},"_tags":["comment","author_karaterobot","story_41663562"],"author":"karaterobot","children":[41680692],"comment_text":"> In private conversations, Mr. Altman has compared the world\u2019s data centers to electricity, according to three people close to the discussions. As the availability of electricity became more widespread, people found better ways of using it. Mr. Altman hoped to do the same with data centers and eventually make A.I. technologies flow like electricity.
Reminds me of a quote I saved from Kevin Kelly's book The Inevitable, which was published in 2016. I saved it because it sounded so absurd at the time. To be clear, it sort of still does sound absurd to me, who remain moderately skeptical about AI (though admittedly less so than in 2016, largely because of the following sentence). What's very serious is how much money and old-fashioned brainpower is going into making this future real. I don't know whether the metaphor was arrived at independently, derives from the same source, or is just an obvious way of thinking about the world when you've drunk the right flavor of Kool-Aid.
> Amid all this activity, of a picture of our AI future is coming into view, and it is not the HAL 9000\u2014a discrete machine animated by a charismatic (yet potentially homicidal) humanlike consciousness\u2014or a Singularitan rapture of superintelligence. The AI on the horizon looks more like Amazon Web Services\u2014cheap, reliable, industrial-grade digital smartness running behind everything, and almost invisible except when it blinks off. This common utility will serve you as much IQ as you want but no more than you need. You\u2019ll simply plug into the grid and get AI as if it was electricity. It will enliven inert objects, much as electricity did more than a century past. Three generations ago, many a tinkerer struck it rich by taking a tool and making an electric version. Take a manual pump; electrify it. Find a hand-wringer washer; electrify it. The entrepreneurs didn\u2019t need to generate the electricity; they brought it from the grid and used it to automate the previously manual. Now everything that we formerly electrified we will cognify. There is almost nothing we can think of that cannot be made new, different, or more valuable by infusing it with some extra IQ. In fact, the business plans of the next 10,000 startups are easy to forecast: Take X and add AI. Find something that can be made better by adding online smartness to it.","created_at":"2024-09-27T23:14:03Z","created_at_i":1727478843,"objectID":"41676462","parent_id":41663562,"story_id":41663562,"story_title":"Behind OpenAI's plan to make A.I. flow like electricity","story_url":"https://www.nytimes.com/2024/09/25/business/openai-plan-electricity.html","updated_at":"2024-09-28T15:30:30Z"}],"hitsPerPage":20,"nbHits":77325,"nbPages":50,"page":0,"params":"query=Cognigy&advancedSyntax=true&analyticsTags=backend","processingTimeMS":22,"processingTimingsMS":{"_request":{"roundTrip":22},"afterFetch":{"format":{"highlighting":1,"total":1},"merge":{"mergeLoop":{"prepareNextHit":4,"total":4},"total":4},"total":4},"fetch":{"query":6,"scanning":10,"total":17},"total":22},"query":"Cognigy","serverTimeMS":24}
![http://www.mactricksandtips.com/wp-content/uploads/2008/03/n...](\"http://www.mactricksandtips.com/wp-content/uploads/2008/03/networked-quartz-composer-patches.png\"){kind=link}
![https://uploads-ssl.webflow.com/5a9d00dba5e9fa00010cb403/5c8...](\"https://uploads-ssl.webflow.com/5a9d00dba5e9fa00010cb403/5c8b82a6e9c13e6c33e03b7a_AM-2019-Screenshot-2.png\"){kind=link}