Recorded on: August 21, 2024
Uncover the various ways to use a RAG solution with PoolParty and GraphDB
We’re familiar with RAG - but which RAG is the right fit?
As organizations increasingly rely on AI-driven information retrieval and generation, traditional Retrieval Augmented Generation (RAG) systems often fall short in production environments. Various solutions have emerged to address this challenge, recognizing that “one size does not fit all” - each organization has unique requirements to tackle this issue within their framework.
Increase efficiency by up to 20% with Semantic RAG (AKA Graph RAG)
This webinar explores an innovative solution: Semantic RAG, which harnesses the power of knowledge graphs, large language models (LLMs), and graph databases to enhance semantic retrieval and generation capabilities.
A key focus will be on LLM-derived knowledge graphs, demonstrating how they can be created and refined with human-in-the-loop processes to achieve greater accuracy and explainability. We'll explore memory constructs in Graph RAG operations, illustrated through practical demonstrations using PoolParty and GraphDB.
We will dive into a real-world case study and explore the pipeline, architecture, and practical applications of Semantic RAG systems. This will showcase valuable insights into the seamless collaboration between PoolParty and GraphDB to deliver a robust Semantic RAG solution. Expert strategies will be shared for maximizing the impact of these innovative technologies by shortening time to insight, providing savvy querying for the untrained, and lowering long term maintenance costs.
Márcia Ferreira, Data and Knowledge Engineer (Technical Consultant) at Semantic Web Company, and Krasimira Bozhonova, Solutions Architect at Ontotext, will walk you through the following aspects and explain the benefits of a Semantic RAG approach:
-
Vector RAG vs. Graph RAG
-
Grounding LLMs with Graph
-
Conventional RAG vs. Semantic RAG
-
Different varieties of Graph RAG
-
Demo of PoolParty Semantic RAG and GraphDB
In the meantime, have a look at our Semantic Retrieval Augmented Generation page to learn more about the capabilities and benefits of utilizing Knowledge Graphs to power your Generative AI experience.
We hope to see you there!
Presented by:
Márcia Ferreira, Semantic Web Company
Márcia R. Ferreira is a Data and Knowledge Engineer (Technical Consultant) at Semantic Web Company. By implementing state-of-the-art semantic systems, Márcia ensures customers' data requirements are met at the highest level. She is interested in integrating LLMs with graph technologies for impactful applications in knowledge-intensive industries.
Krasimira Bozhanova, Ontotext
Krasimira Bozhanova is a Solutions Architect at Ontotext, starting her way as a developer and contributing to many of the company's significant client projects. In recent years she has been a part of the AI in Action Ontotext initiative which puts Ontotext products at the heart of various AI applications.
Large Language Models (LLMs) are a hot topic, discussed from many different angles. There is a lot of excitement about what they can do, but also a lot of concern about how they do it.
It's one thing to let generative AI write a poem or the description of a talk (not the case here), but can we trust it to give us the right answers to critical questions or make the right decisions? The clear disadvantages of this approach are the lack of explainability of the results, consistency, and accuracy of responses, the need for large amounts of high-quality data, and the potential for bias in the content generated by the models. One possible answer to these challenges is to combine these generative technologies with knowledge graphs that can help explain the results and support the generation of high-quality data to train the models. The combination of generative AI and symbolic AI can ultimately lead to eXplainable AI (XAI), which can help us in our daily work.
We will show the benefits of combining both technologies, how knowledge graphs and intelligent structured source content can help to create better and more explainable results with generative technologies, and how generative technologies can help to create better knowledge graphs. This combination will lead to the creation of intelligent content that enables deeper understanding and informed decision making.
Etteplan and Semantic Web Company have developed unparalleled experiences and capabilities in creating, developing, and maintaining consistent (technical) knowledge bases for many companies across multiple industries and markets - placing them in a unique position to leverage the true potential of Large Language Models.
Presented by:
Helmut Nagy, Semantic Web
Helmut is Chief Operating Officer of the Semantic Web Company (SWC). In his role as COO he designs and implements customer-centric processes and plans the strategic development of the PoolParty Semantic Suite together with the involved teams based on the SWC business strategy.
Eric Tengstrand, Etteplan
With over 10 years of experience in leading and developing service solutions for different customers within the manufacturing industry, Eric is currently the Vice President Global Service Solutions in Technical Communication Solutions at Etteplan.
Etteplan is a rapidly growing Technology Service company specializing in software and embedded solutions, engineering solutions, and technical communication solutions. We are a forerunner in the engineering industry and we differentiate ourselves by the wide-ranging competence of our experts. Our customers include the world’s leading companies in the manufacturing industry. We help them to create a better world through engineering, innovation, and digitalization. The company currently has over 4,000 professionals globally of which 850 are in Technical Communication Solutions.
Eric’s mission is to help Etteplan’s customers improve their operational performance and achieve their business goals through innovative and cost-effective managed services.
Eric is responsible for the strategic direction, offering development and solutions sales leading a team of professionals across 4 countries.
Stefan Gentz is the Senior Worldwide Evangelist for Technical Communication at Adobe, headquartered in San Jose, California, USA. Stefan’s mission is to inspire enterprises and technical writers around the world and show how to create compelling technical communication content with the Adobe Technical Communication tools. He is also a certified Quality Management Professional (TÜV), ISO 9001 / EN 15038 auditor, ISO 31000 Risk Management expert, and Six Sigma Champion.
As a sought-after keynote speaker and moderator at conferences around the world, he travels around the globe half of the year. Besides that, he has been the European Ambassador for the Globalization and Localization Association (GALA) for many years, a member of the tekom Conference Advisory Board for several years, and a founding member of the tekom 
Saibal Bhattacharjee has been with Adobe for more than 12 years now and is currently working as Director of Product Marketing & Business Strategy in the Digital Advertising, Learning & Publishing business unit.
Saibal is in charge of GTM and Business Strategy for a $140MN+ product portfolio in Adobe – ranging from market-leading cloud-native component content management system (CCMS), advertising & subscription monetization products for connected multiscreen TV platforms, content authoring, and publishing desktop apps, to print & scan solutions that power Office and Commercial printers, etc.
Saibal drives the products’ positioning and messaging, brand strategy and communications, marketing campaigns and demand generation (growth and retention), partner ecosystem and new market development, customer education, business insights, market research, and analyst relations, sales enablement, and field marketing efforts.
LinkedIn: 
Stilo showcases a comprehensive, end-to-end process for converting pharmaceutical documentation, beginning with the source DOCX Word file. This journey encompasses the creation of automation rules, execution of the conversion, meticulous quality assurance, and, ultimately, the seamless publication of the final output through Oxygen. Explore how Stilo's Migrate effortlessly addresses prevalent conversion challenges, including the precise identification and generation of product name keys, resolution of table alignment discrepancies, and an array of other crucial tasks. Dive deep into the intricacies of transforming your industry-specific content from its initial state, ensuring a flawless and efficient transition.
Presented by: TJ Dhaliwal, Stilo
TJ Dhaliwal is a Technical Sales Product Specialist at Stilo, helping customers uncover solutions to their content conversion needs. With a solid technical background in engineering and a passion for public speaking, he’s successfully bridged the gaps between project stakeholders. As a problem solver, TJ has helped several enterprise-level companies convert tens of thousands of pages of their content to DITA.
Dawn Stevens, President, Comtech Services and Director of CIDM has 28 years of practical experience in virtually every role within a documentation and training department, including project management, instructional design, writing, editing, and multimedia programming. Dawn is the perfect advisor to identify and remove the challenges you face in producing usable, technical information and training content. With both engineering and technical communication degrees, Dawn combines her solid technical foundation with strong writing and design skills to lead our team of consultants and specialists in providing the expertise you need.
Jacob Brennan is a Technical Sales Product Specialist at Stilo. He works with Stilo Migrate customers and aids in converting their legacy content to DITA. Jacob is a recent graduate from the University of Ottawa holding a BASc. in Mechanical Engineering and a BSc. in Computing Technology. 
Klaus Fleischmann studied translation and IT in Vienna, holds an MA in Conference Interpreting from Monterey, California, and a MAS in Technical Communication from Krems, Austria. In 1996, he founded Austria-based Kaleidoscope, a company implementing content, translation, and terminology management processes for internationally active companies. Kaleidoscope develops online collaboration software for enterprise-level terminology workflow, translator query management, in-country review etc., making the translation quality process comprehensible and strategically manageable. In 2007, he became CEO of Austria´s leading LSP, Eurocom Translation Services. Always active in the industry, Klaus got voted into the Gala Board of Directors in 2015 and 2017. 
Tristan is Product Director at DeltaXML, a technology company with world-leading software products for the management of change in structured content. He has a deep understanding of DeltaXML’s product suite and loves to help customers create extra value in their content using change management. Tristan is a father to three daughters, a movie lover, and a keen runner.
Fabrice Lacroix is a serial entrepreneur and a technology pioneer. He has been working for 25 years on the development of innovative solutions around search technology, content enrichment and AI. He is the founder of Fluid Topics, the leading Content Delivery Platform that reinvents how users search, read and interact with technical documentation. 
While product cycles are getting shorter in all industries, down to daily releases for SaaS vendors, technical documentation teams have to keep up with the pace. Information needs to be written faster and delivered in sync.
Join Fabrice Lacroix, CEO and founder of Fluid Topics, and Gaspard Bébié-Valérian, certified technical writer and functional consultant, as they untangle the complex process of continuous content delivery and unveil new methods and processes to embrace the new landscape of documentation. They will share their insights and experience as they explore the collaborative potential between SMEs and tech writers opened by Docs as Code, examine the tooling behind a continuous documentation flow, and showcase the example of a robot writing release notes.
Presented by Fabrice Lacroix & Gaspard Bébié-Valérian:
Fabrice Lacroix is a serial entrepreneur and a technology pioneer. He has been working for 25 years on the development of innovative solutions around search technology, content enrichment and AI. He is the founder of Fluid Topics, the leading Content Delivery Platform that reinvents how users search, read and interact with technical documentation.
At the crossroads of engineering and social sciences, Gaspard enjoys exploring the breadth of technologies involved in technical documentation, including structured documentation, CCMS, and collaborative paradigms like docs as code. Gaspard works as Functional Consultant at Antidot helping clients get the most out of Fluid Topics. 
Helen joined Stilo as a technical editor. She now works closely with Stilo Migrate customers, helping them to analyze their legacy content and configure appropriate mapping rules. She also provides Migrate customer training and support. Helen has helped Migrate customers to convert millions of pages of content to DITA and custom XML. Helen holds a Bachelor of Arts in English from St. Francis Xavier University in Antigonish, Nova Scotia, and has pursued graduate studies at Queen’s University in Kingston, Ontario. Email: 
Christian Lang, Technical Consultant, has a wide range of experience and interests in the language field, proven by his degree in Japanese studies and translation. He first became involved with terminology management as a freelance translator for the European Patent Office. Since then, it has become one of his hobbies, as has research in the field of NLP on topics such as machine translation, automatic term extraction, and concept maps.
