# Thought leadership

Ideas and insights from our team of experts working at the cutting-edge of generative AI development.

## [The living brain of the enterprise](/content/engineering/enterprise-as-living-brain/index.html)

**Matan-Paul Shetrit**

## [The orchestration graph](/content/engineering/orchestration-graph/index.html)

**Matan-Paul Shetrit**

## [Everyone is a manager now](/content/engineering/employee-into-manager/index.html)

**Matan-Paul Shetrit**

## [Supervising the synthetic workforce: Observability for AI agents requires managers, not metrics](/content/engineering/supervising-synthetic-workforce/index.html)

**Matan-Paul Shetrit**

## [Anyone can build software now — and it’s causing hell for developers](/content/engineering/agent-development-lifecycle/index.html)

**May Habib**

## [Self-evolving models, not test-time compute methods, are the future of AI](https://www.linkedin.com/posts/waseemalshikh_test-time-compute-methods-like-r1-o1-o3-activity-7291269564900782080-E1Ep?utm_source=share&utm_medium=member_desktop)

## [The incentives of model innovation](https://www.linkedin.com/posts/waseemalshikh_lots-of-hype-about-deepseek-and-its-r1-llm-activity-7288684074699816961-1UEy?utm_source=share&utm_medium=member_desktop)

## [A breakdown of LLM agent types](https://www.linkedin.com/posts/waseemalshikh_as-im-sure-everyone-has-been-shouting-from-activity-7283879651062558720-37Oa?utm_source=share&utm_medium=member_desktop)

## [Synthetic data: Busting the myths holding back enterprise AI progress](/content/engineering/synthetic-data-myths-vs-facts/index.html)

## [Navigating the challenges of generative AI and "vendor lock-In" in enterprises](/content/engineering/vendor-lock-in-generative-ai/index.html)

## [The three keys to escaping AI POC purgatory](/content/engineering/escaping-ai-poc-purgatory/index.html)

## [Why AI-native enterprise apps are the business brain of the future](/content/engineering/ai-native-apps/index.html)

## [Vector database vs. graph database](/content/engineering/vector-database-vs-graph-database/index.html)

## [RAG vector database explained](/content/engineering/rag-vector-database/index.html)

## [The new paradigm: AI-native enterprise applications](https://www.linkedin.com/posts/waseemalshikh_as-a-cto-with-over-a-decade-of-experience-activity-7211435753275277312-2ZOw)

## [Short transformers: easily prune redundant LLM layers](https://www.linkedin.com/posts/melisa-russak-5b7987145_short-transformers-easily-prune-redundant-activity-7201220721677635584-lWcg)

## [The rise of domain-specific Large Language Models](https://www.linkedin.com/posts/waseemalshikh_in-the-rapidly-evolving-landscape-of-aillm-activity-7210758077128568833-wkfj)

## [Q&A with Knowledge Graphs](https://www.linkedin.com/posts/waseemalshikh_fid-fid-knowledgegraphs-activity-7193474643377221632-sTc9)

## [Insights from LLM Control Theory](https://www.linkedin.com/posts/waseemalshikh_gpt4-palmyra-llama-activity-7192582597774880768-pfeB)

## [OmniACT: a groundbreaking dataset and benchmark for autonomous agents](https://www.linkedin.com/posts/waseemalshikh_omniact-palmyra-ai-activity-7170402655897378818-hHqL)

## [Why we chose DPO over RHLF](https://www.linkedin.com/posts/waseemalshikh_dpo-dss-rlhf-activity-7157908439803793408-FUdo)

## [LLMs: Skills, behaviors, and knowledge](https://www.linkedin.com/posts/waseemalshikh_llms-skills-llms-activity-7154994885786288128-XmiY)