Palantir Platform

My main use case for Palantir Foundry  involves end-to-end solutions, starting from data ingestion and transformation. I also use AIP logics for data transformation and ontology management.

In my day-to-day work, I create or modify back-end functions using TypeScript to create action types under the ontologies to add new rows, which means adding new objects or updating existing ones. I also use AIP logics to accomplish the same tasks.

While using AIP logics, I work with machine learning LLM prompts to get results based on text from various unstructured files, such as PDFs. I use the LLM to extract data for analysis in workshops for Contour.

Palantir Foundry  offers many good features, including ontology management and user-friendliness. I find AIP Logic LLMs quite useful for small tasks, such as finding patterns or trends between two data sets. I frequently use the Object Explorer, now called Site Insight, which is useful for comparing data and performing sanity checks.

The accuracy and reliability of Palantir Foundry's output are quite good. During analysis, we find its accuracy satisfactory.

Palantir Foundry has positively impacted my organization by providing good support from Palantir teams, facilitating the development of many new solutions, building our UI and web applications, and significantly enhancing our productivity. Overall, we enjoy using it.

When issues arise, I often have to rely on Palantir support team rather than solving them myself, as some team features require heavy reliance on their support due to a lack of comprehensive documentation for us to configure things on our end.

Although there are enough tutorials and Palantir study materials available, we still face many struggles when working on projects within Palantir and often need to consult the support team, particularly regarding synchronization and visibility on back-end processes.

I would rate Palantir Foundry a seven because it needs to be more open. As a developer, I find the limited documentation and less resource availability restrictive compared to other options such as AWS .

My rating of seven is influenced by our heavy dependence on Palantir support engineering team to resolve issues that we cannot. The flexibility of some features, such as ontologies, could improve.

I have been using Palantir Foundry for the last nine months, since September.

Palantir Foundry is stable.

Its scalability is good.

Customer support is good, and we interact frequently with the support team.

I would rate customer support an eight.

My advice for others looking into using Palantir Foundry is to have customer support at the beginning, as venturing into it without much knowledge could be challenging.

I did not previously use a different solution before Palantir Foundry.

I see that there are metrics showing we have saved money using Palantir's features, indicating a return on investment and improved efficiency after implementing Palantir solutions.

We prefer developing our functions and modules in Palantir rather than using outside sources, leading to the recent development of two or three models in Palantir that previously relied on outside sources. This shift saves costs and improves the stability and consistency of our services.

The code repository is critical for us. Our entire back-end relies on it, and everything we do or write in functions is stored there, making us completely dependent on that repository.

I believe the governance and security of Palantir Foundry are quite secure, which played a key role in our choice of Palantir. While its AI capabilities are not as advanced as other leading tools such as ChatGPT, it is still useful and shows improvement, especially with the LLM feature on AIP Logics.

I rate Palantir Foundry a seven overall.

My main use case for Palantir Foundry  is document processing, taking large volumes of documents, scanning and parsing them, and using AIP to generate insights and applications based on that work.

I continue to expand the functionality and leverage ontology to provide even more accurate and business-friendly metrics.

Palantir Foundry  offers excellent features that make it very user-friendly with quick setup, allowing me to build applications rapidly and then extend those into production applications.

My day-to-day work involves working independently to stand up some applications or collaborating with engineers to conduct quick prototyping.

Palantir Foundry has positively impacted my organization by centralizing the way we build applications and standardizing business metric reporting. This centralization and standardized reporting have created more alignment across my team and the organization, greater productivity between teams, and quicker cycle times into production.

To improve Palantir Foundry, I would appreciate continued updates to the documentation and more tooltips, auto-correct features, or auto-validation on certain types of information.

I have been using Palantir Foundry for approximately two years.

Palantir Foundry is stable.

Palantir Foundry's scalability is very high, given that once the ontology and core applications are set up, it is easy to deploy those and manage the costs effectively.

Palantir Foundry's customer support is great.

My experience with pricing, setup cost, and licensing has been relatively easy, and the pricing made sense given the nature and scope of our work.

I have seen return on investment with time saved on overall development time from initial build to deployment, and overall engineering productivity has increased.

My advice to others looking into using Palantir Foundry is to ensure that their teams are aware of how to use the ontology most effectively and have a general understanding of best practices around AI application setup. I would rate my overall experience with Palantir Foundry as an 8 out of 10.

I have been working on projects related to company products, particularly streamlining processes that hold company data. This involves finding pipelines, creating forecasts, and determining scopes and measures for mapping things in a single umbrella where all resources can be mapped through.

Apart from that, I have worked on products for the Japan government. I cannot disclose what the specific product is, but it is more into the risk side. There is a particular risk measure related to earthquake-related things, specifically finding the seismic intensity of earthquakes. This involves figuring out points using coordinates and plotting them into a map to have an understanding of earthquake patterns and complexity. Earthquakes follow certain patterns, and this is exactly affecting warehouses, industries, and factories located near bay areas. I particularly worked on plotting the magnitude of earthquakes greater than 7.5, which is considered as the red zone. Japan is one of the countries with the maximum number of earthquakes. To accomplish this, it creates auto-alerts or something similar. Based on that same use case scenario, we created applications and pipelines where we can get day-to-day data. It can create a pre-warning and let the user be aware of similar patterns if they are happening, because they have a certain metric system of how they calculate patterns and risks. This is a complete system being built on Palantir Foundry .

Palantir Foundry  is quite secure in terms of where we can access the data from different regions. I have used it across multiple teams across the globe, with my team diversified in Japan, Philippines, India, and China. The security is top-notch, and we have shared services across all divisions. Furthermore, by integrating various resources under a single umbrella, Palantir Foundry plays a significant role in streamlining processes and data management for various projects.

I believe there are certain ways Palantir Foundry could get improvements. Palantir Foundry is majorly in the data engineering domain where you write down code and intake data. It could be simplified into action types where certain things can be simplified rather than being complicated. Under the hood, if someone works with it, they actually need to know the know-hows in depth about things. Otherwise, it will be really difficult for them to work on.

Given the scenario, if a person from an application development background, with less data engineering domain knowledge, comes to work with Palantir Foundry, they might find it a little tricky because things are not usually the same. They have their own jargons and their own sort of understanding. They need to understand the ontology of how Palantir Foundry works. There are certain things under the hood that could be simplified, but it is made in its own way.

I have been using Palantir Foundry for more than one and a half years, or about eighteen months or more.

Regarding stability, there are certain things I need to mention. When you create pipelines and do a build, without knowing what you are doing, there are chances where a lot of failure cases occur. There are a lot of failed builds that happen. If that could be checked at the earliest stage, it would be beneficial, not only for the developers who are working, but also for the person or the company who are putting in the time. When we develop or do the build, there are many chances that it fails. It could fail due to multi-existence of data, because we really do not check while we work. There are points when you are continuously working, so you really do not have to look into the ontology or the objects majorly, but you have to develop certain things. At that point in time, if there could be certain suggestions given by the system, that would be really helpful.

I have never contacted the technical support or customer support. Most of the things when we had an issue, my team and I particularly tried to troubleshoot using the documentation majorly, because they are quite pretty clean and clear about the documentation.

When I started with Palantir Foundry, the initial deployment was not that great. I was not able to figure out certain things, because there are certain coding patterns that need to be followed in order to make it run. Palantir Foundry does not have the usual patterns that need to be followed.

Palantir Foundry has its own set of code and coding practices and coding standards you have to follow. There are options where you can implement certain coding methodologies like Python or PySpark for doing transformations. You can implement that, but there are certain things that people must be aware about. Without knowing that, if you just jump down to Palantir Foundry, it will be a little bit tricky initially. When I started working with it, having an AI agent or a chatbot was not so expected in applications. Later, I feel there is an application bot where if you ask questions, it will let you know certain things or it can point you out to certain links which you can follow through and get the information.

I am not aware about the pricing at all. I am currently working with one of the reputed companies in Japan. Regarding the pricing and the domains, everything is being completely taken care of by the finance unit. They have certain things for the enterprise editions. There could be options for personal use, but I have never really checked out because I have been working on a company laptop where everything is being integrated through a single SSO .

I did not get an opportunity regarding the alternatives to Palantir Foundry. There were certain limitations because of something that comes with the billing structure. My company is already partnering with Palantir, where certain billing cycles are expected monthly. It is done monthly based on the number of users. There are options where Palantir Foundry is available with Azure , but we have never really checked out those options. Considering Palantir Foundry, it is always trustworthy.

My company has a partnership with Palantir, but basically, I am just a customer. Most of the things we design and develop are being taken care of by us. I have never used any service beyond that, even if it exists. I can give around eight out of ten on a scale from one to ten for Palantir Foundry overall for everything.