GetFocus Feature Guide

1 exploring and scouting technologies 1 exploring and scouting technologies begin by scouting technologies to find all the emerging technologies competing in your field of choice this helps set the foundation for a comprehensive technology forecast technology scouting technology scouting ¹ ¹ this chat feature gi v v es results for all the technologies in a given area for the search query the reasonig² reasonig² offers greater accuracy and excel at handling complex tasks at the expense of speed use the reasoning button when accuracy is more important than speed the results ³ ³ from technology scouting outputs list of all the relevant technologies grouped as industry standard technologies, emerging technologies and novel/niche/early stage technologies ensuring coverage from early stage to mature technologies some example queries for technology scouting basic "battery chemistries" with context "battery chemistries for electric vehicles" detailed "emerging high energy density battery chemistries for next gen electric vehicles" ask me anything ask me anything ⁴ ⁴ this chat feature scans the internal knowledge base of our llm and allows you to create custom prompts if you start typing in the text box without selecting a chat option, the "ask me anything" feature will be automatically triggered in the chat, a large language model (llm) generates follow up question ⁵ to help you dive deeper into and better understand specific technologies summarize technologies (optional) provides concise summaries of the scouted technologies evaluate for application area (optional) analyzes the relevance of the scouted technologies for a specific application area you provide you can then instruct the model with an additional prompt to output results in a list or a table format you can then instruct the model with an additional prompt to output results in a list note all the chats are saved and can be accessed on the left side of the tool ⁶ 2 evaluating technologies 2 evaluating technologies post scouting, evaluate the emerging technologies to assess their suitability for your specific use case leverage "ask me anything" to define custom criteria and evaluate technologies this approach offers flexibility to use metrics that best suit your requirements here is an example ¹ for technology evaluation using ask me anything best practices for prompting best practices for prompting persona based prompts have traditionally shaped the style or tone of responses, and remain valuable when you want to influence language style, tone, or domain specific flavor however, newer reasoning capable llms often incorporate internal mechanisms that handle logical inference and multi‑step thinking in such models, explicit personas or prolonged instructions can impair performance instead,it’s better to keep prompts minimal and precise here are some examples persona based prompt persona based prompt "you are battery chemistry expert you know everything there is to know about relevant technologies regarding battery chemistry i will give you a list of battery technologies, as an expert evaluate the technologies based on given criteria " revised prompt revised prompt "evaluate the following battery technologies based on given criteria \[criteria]" in this example, our focus is on battery technologies, thus the evaluation will be based on cost efficiency, energy density, cycle life, safety, and sustainability be sure to define how scoring should work (for example, using numeric values, yes/no options, etc ) you can ask the tool to explain the reasoning behind each score—for instance, “ after each criterion, add a column with the reasoning for the score ” finally, feel free to add " present the evaluation results in a clear table format " 3 other chat features 3 other chat features while the first four chat options are used for scouting and evaluation of technologies, the other chat functions are useful for other tasks new technology application finder new technology application finder this chat feature finds new application and commercial opportunities for a given technology solutions from other industries solutions from other industries this chat feature explores solutions to technology problems in other industries to find inspiration for solving technological problems solutions from nature solutions from nature this chat feature explores solutions in nature to a technological problem, which is helpful for identifying possible solutions through biomimicry after initiating the chat you can use follow up questions to dive deeper and continue the conversation visit help center > getting started with odin > how to use chat? https //docs getfocus eu/how to use chat for more information 4 patent landscape 4 patent landscape after scouting and evaluation, narrow down to top scoring technologies for your use case next, map out their patent landscapes to identify improvement rates technology search technology search technology¹ search helps you find all patent families associated with your search term(s) use technology search when you are looking to create a landscape of patents associated with a certain topic selecting the appropriate search algorithm ² is crucial to achieving the desired balance between precision and recall agentic search agentic search the default algorithm, best for creating a highly relevant patent landscape with a focus on precision if results are too narrow, reduce the similarity filter to include more patents smart search smart search the generative ai transforms your keyword input into a context rich, four sentence technical description these enhanced queries are ideal for vector search, leading to improved results exact search exact search offers complete control over search parameters and is ideal for precise targeting and custom queries to meet specific objectives organization search organization search organization³ search allows you to find all patent families associated with one or more organizations it is especially useful when you want to explore the complete patent portfolio of a specific company or compare multiple organizations simultaneously publication number search publication number search publication number⁴ search allows you to find specific patents or patent families using their publication numbers use this feature when you need to retrieve a particular patent or search for multiple publication numbers simultaneously similar inventions search similar inventions search similar inventions⁵ search allows you to find patents that are closely related to a known invention by using its publication number this feature is ideal for discovering similar innovations and exploring adjacent technologies you can view how the agentic search and smart search enriched your original input by clicking the info icon ⁶ on the right side of the search title you can also copy the smart query, edit it to suit your technical requirements, and then paste it into the search box finally, select the exact search option in the technology search to run your customized query visit help center > searching with odin https //docs getfocus eu/searching with odin for detailed guidance llm filters llm filters llm filters feature enables you to find patent families based on complex queries like "find patents related to pyrolysis, a thermochemical decomposition process operating under an inert atmosphere at temperatures between 400°c and 650°c" or you may also filter a dataset simply using natural language like “ search for all patents that discuss cryogenic carbon capture ” below are some universal best practices best practices that help to get relevant results make the task objective clear make the task objective clear use strong, directive phrasing like “explicitly,” “specifically,” “focus on,” or “discuss” to clarify objectives avoid vague verbs such as “mention”, since a patent might reference a technology without it being the core focus use logical connectors carefully use logical connectors carefully pay close attention when structuring compound queries for example, “application to fault detection or maintenance planning” will return broader results than “application to fault detection and maintenance planning ” this distinction can significantly impact the output include exclusion conditions when needed include exclusion conditions when needed consider adding constraints like “if not all conditions are met, exclude the patent ” this enforces strict filtering logic and helps ensure results align precisely follow these steps to filter dataset using llm filters click on create a new llm filter ¹ enter your query in the text box² and click create llm filter ³ after the filter is ready click on apply filter ⁴ to create another llm filter, click on see all llm filters ⁵ and create new ⁶ to switch between different filters you created, select the filter and click on apply note this process is time consuming and may take 3 60 minutes depending on the size of the dataset llm filter runs in the background, so you can multitask until it's done processing your data llm filter visit help center > searching with odin > how to filter search results? https //docs getfocus eu/how to filter search results for detailed guidance keywords keywords filter filter after performing a smart search, you will likely find a large number of patents our ai search model helps you get in the right direction, but filtering is crucial for accuracy follow the steps below to refine the set and create a good representative patent landscape to derive a precise improvement rate including keywords¹ use terms that are generally used to define the technology or relate to the technology eg excluding keywords remove terms that may pollute the landscape for the technology sort patents² in ascending order of relevance to check if the least relevant still pertain to your technology if the least relevant results are still relevant, you can be sure that the rest of your dataset is also accurate visit help center > searching with odin > how to filter search results? https //docs getfocus eu/how to filter search results for detailed guidance after sorting by ascending relevance, titles and abstracts might not always show a patent's relevance use ai summary³ or chat with invention⁴ with your queries for clarity if most of the least similar patents remain relevant, the landscape is accurate enough if the first page of ascending relevance doesn't match your field, try filtering with exclude keywords⁵ from non relevant patents eg lithium ion, sodium increasing similarity⁶ by 0 5 1% per step for better results after ensuring the patent landscape accurately reflects the intended technology, save⁷ the set and repeat these steps for other technologies to compare their technology improvement rates remember to “ save ” the dataset and always “ save changes ” when using filters 5 technology improvement rates 5 technology improvement rates once your technology sets are ready, compare their improvement rates by following these steps select your project folder click on the compare¹ link at the top left select your desired technologies by checking the boxes ² click the compare selected³ technology improvement rate (tir)⁴ is a comparative metric, not an absolute indicator there's no given improvement rate that signifies whether a technology is advancing rapidly you must compare it to competing technologies to gauge whether it is improving faster than the competition the 'delta' between improvement rates reveals which technology is likely to dominate historically, the fastest improving technology among competitors tends to win for a reliable dataset, make sure the patent set does not include less than 50 patents the more the better visit help center > searching with odin > managing projects https //docs getfocus eu/managing projects for detailed guidance 6 forecasting 6 forecasting after comparing technologies, click on select charts¹ and select tir forecast to view the forecasted technological performance chart here we predict when competing technologies might take the lead in the future you'll need to input the starting values for current costs and performance of the technologies, which can be difficult to obtain since the tool creates charts from your data, ensure the inputs are accurate—otherwise, you risk getting unreliable results be careful what you put in, garbage in = garbage out here is how the forecasting tool works first click on settings² we use the latest known improvement rates of the technologies, which we obtained in previous steps we use the approximate current costs³ of the technologies for a particular performance metric (in our example, it is $/ kwh) interpreting the forecasting curve graphene batteries, with an improvement rate of 67 5%, are projected to surpass solid state lithium batteries by 2033, despite their high current cost of $1000 per kwh sodium ion batteries will continue to be the most cost effective option for the foreseeable future due to their significantly lower starting costs, even though their improvement rate is lower please note this is only an example, the improvement rates and costs in the example are arbitrary values selected for a simple explanation 7 deep dives 7 deep dives once you've identified the best technology, gain detailed insights by chatting with individual patents or a set of patents ai summary ai summary click the ai summary¹ button to get a concise overview of any patent it extracts and displays rewritten patent abstract for maximal clarity technologies used in the invention advantages of the invention application areas chat with invention chat with invention use the chat with invention² feature to q\&a a specific patent via natural language, powered by a large language model (llm) you can also select the reasoning option for more complex questions and the model returns a precise answer to your question responses are derived solely from the patent text our analysts frequently use it for open ended questions "summarize this invention to an engineer" for simple technical understanding specific information "what are the electrode materials mentioned in this invention?" ³ claims inquiry "what are the dependent/independent claims in the invention?" chat with set chat with set the chat with set⁶ function allows you to analyze up to 1000 patents using natural language, powered by a large language model (llm) you can ask your dataset any question, and the llm will analyze and provide answers typical use cases for "chat with set" finding the latest trends⁷ what are the key trends in this domain over the past 5 years? what are the topics being discussed in this dataset and how many inventions belong to each topic? comparing inventions what are the differences between patent us 20110052986 a1 and patent cn 117525548 a? how did this technological domain develop? compare 5 years ago with the current year and explain the differences save "chat with set" click on save⁸ button to store your current chat conversation visit help center > chat with set https //docs getfocus eu/chat with set for detailed guidance click on overview⁹ for better understanding of organizational landscape for key technologies it helps you to identify key players recognize all organizations (public and private) innovating in key technologies identify geographical trends find out which regions are patenting more or less in your technology area analyze activity evaluate organizational activity by assessing patent volume and quality use "chat with set" for organization and portfolio analysis what are the top 3 startups in this domain? provide a summary of their most important inventions compare the portfolios of sumitomo industries and yuasa battery inc and explain the main differences 8 tracking workflows 8 tracking workflows tracking workflows helps to monitor your technological domains and allows you to keep track of and analyze relevant new inventions automatically initiate workflow go to tracking ¹ > add tracking workflow ² add name and description of your workflow ³ set up trigger node choose a schedule daily, weekly, monthly, or yearly ⁴ add action nodes depending on your need select an action node ⁵ llm agent applies a provided prompt to every new patent since the last run eg "describe in 5 sentences what is novel about the latest patent " get latest patents fetches all new patents added since the last run summarize patents generates summaries for each new patent since the last run click on add node after each step to add them to your workflow configure notification node select "send email" ⁶ input recipient email addresses ⁷ test the workflow click test workflow ⁸ to check functionality finalize and activate save your workflow it's now scheduled to run automatically ⁹ visit help center > tracking workflows https //docs getfocus eu/tracking workflows for detailed information