News & Intel
LiveDaily AI-curated intelligence on Palantir Foundry, Ontology, AIP, Apollo, contracts, and community feedback. Updated automatically via GitHub Actions every day at 7 AM UTC.
Texas Freight Corridor Pressure Monitor — Corridor-Week Anomaly Detection in Foundry & AIP
I built a Foundry-native application that turns public Texas transportation data into an operational corridor pressure monitor. For any selected week, it answers one question: Which Texas freight corridors are behaving normally, running above normal, or showing below-normal/disrupted movement compared to their own historical baseline? Stack: Pipeline Builder → Ontology → Workshop → AIP Generated Content Validation event: Winter Storm Uri, week of 2021-02-15 Object grain: one corridor during one week The system is deliberately scoped as a corridor pressure anomaly monitor . It is not a diesel demand, truck-demand, or fuel-consumption model. The goal was to demonstrate a complete, inspectable Foundry implementation pattern end to end. Why this problem Freight corridors are operational systems. When movement across a major corridor deviates from its own normal pattern, that deviation can reflect weather disruption, logistics stress, demand shifts, or r
Agentic Workflows in Foundry: Questions, Answered
We ran a webinar last week on building agentic workflows in Foundry – walking through the tooling you can reach for (Chatbot Studio, AIP Logic, Agent Flow etc.) and how to wire them together. The interesting part wasn’t the tooling overview. It was the questions that came after, once people started thinking about what happens when these agents actually run in production. How does an agent behave when it’s writing data, not just returning an answer? Where is it even running if it’s exposed as a function? What breaks when a workflow runs longer than five minutes? How do you keep sessions and permissions sane? Those are the questions we’re answering below. They came straight from the session, and they’re the ones worth getting right before you publish. What is the execution context of these agents, and how do they affect side effects? An agent is not necessarily a pure function that receives an input and returns a complete answer. The function-shaped interface is
Palantir Durable Agents
HN Discussion — 0 comments · 3 points.
When will GPT-5.6 Sol/Terra/Luna, Grok 4.5, and Claude Fable 5 be available in AIP?
Hi Palantir team! What is the expected timeline for adding the following models to AIP’s Language Model Service? GPT-5.6 Sol, Terra, and Luna Grok 4.5 Claude Fable 5 For GPT-5.6 and Grok 4.5, when should we expect them to become available? For Claude Fable 5, which has been generally available since June 9 but is still not listed as a supported AIP model, is it planned for addition? Anthropic states that Fable 5 requires 30-day data retention and is not available under ZDR. Does that requirement affect whether or when it will become available in AIP? We’re interested in using these models across AIP workflows and comparing their performance for our use cases through AIP Evals. Could you also share whether support will extend across AIP Logic, Functions, transforms, agents, and Evals? Thanks! 2 posts - 2 participants Read full topic
Cannot register legacy uppercase-named same-region S3 bucket
Goal I’m trying to add a same-region S3 bucket with READ ONLY access to my enrollment’s S3 bucket policies allow-list. Background In the connection settings of my Data Connection source, it states: S3 buckets in the same AWS region where your Foundry instance is hosted must also be added to the S3 bucket policies allow-list on the network egress page in Control Panel. Control Panel: Network egress → S3 bucket policies → AWS S3 Bucket Policies I navigated there and attempted to add an AWS S3 Bucket Policy, with it correctly detecting the AWS bucket region: us-east-1 . But when I click save, nothing happens and no error message is visible on the frontend. I believe this is because the bucket has a legacy uppercase name, made before March 2018; it accepts test1 but silently rejects TEST1 — both validated as us-east-1 before I clicked Save. Questions Is there a supported way to register a same-region S3 bucket with a legacy uppercase name? Why is no e
Governance for Security-Sensitive Controls
Hi One area where I think the platform could be strengthened is around the lifecycle management and governance of security-sensitive administrative decisions . Today, in most stacks, certain capabilities such as exports, creating third-party application registries, egress policies that allow integration with external systems are managed by admins and used based on an organization’s risk appetite. However, what seems to be missing is ongoing checking after those decisions have been made. Here’s a few examples that come to my mind: An export is a risk and may have been approved based on a specific assessment and business justification. Over time, the data connection configuration (agents, or egress proxy configuration) could change - or even the external source system configuration - changing the risk profile and allowing the original approval to be used in ways that were never part of the initial assessment. Exports with marking validations in REST API data connections
Matrix for TSV2 Function Timeouts
Other posts have touched on parts of this topic separately. AIP FDE, AIP Assist, and my own documentation reviews cannot come up with definitive answers. I am fine with running things through Automate when needed, but I would like to definitively know when that needs to occur and if AIP Logic Functions vs TSV2 code functions in deployed mode will be needed. An answer here would be great, but ideally, this matrix would be published in the documentation so AI FDE can know it. Summary I’m looking for a comprehensive reference that documents the timeout limits for Foundry functions across all execution modes, languages, and calling contexts. This information is partially surfaced in the deployed function configuration panel but is not available in the official documentation, making it difficult to architect solutions that involve long-running function calls. Background I’m working with a TypeScript v2 repository that contains a mix of lightweight query functions and hea
LLM Wiki Architecture in Foundry?
Hi everyone, Andrej Karpathy’s “ LLM Wiki ” pattern has been a hot topic in the agent field. We’re building several agents in Foundry and think this pattern could work well, but Foundry doesn’t have an easy way to give an agent a raw markdown filesystem it can freely create and edit. Notepad seems like the closest option, but feels limited. I’m guessing not many people have actually tried this yet, so I’d love to hear your thoughts even without hands-on experience: If you had to build the “wiki” layer in Foundry, what would you reach for ? Notepad, a dataset-backed store, Media sets, something else? How would you handle agent write access and validation before edits get committed? Curious how you’d approach this. Thanks a ton! 1 post - 1 participant Read full topic
Piq Energy Raises Oversubscribed $5 Million Seed Round to Help Unlock North America's Grid Bottleneck
As AI, electrification and industrial growth drive unprecedented electricity demand, Piq's agentic grid planning platform helps utilities and developers connect projects to the grid faster by automating complex engineering workflows. As AI, electrification an…
foundry worker + agent proxy architecture: webSocket usage and session management
Since Agent Worker is gradually being phased out, we are considering migrating to Foundry Worker + Agent Proxy. With Foundry Worker + Agent Proxy, WebSockets are used between the Agent and Foundry. However, depending on corporate policies, WebSockets may not be permitted. At the very least, we want to clarify when sessions time out. Could anyone please advise us on this? 1 post - 1 participant Read full topic
Tool calling is broken for Codex with LLM proxies
The LLM proxy for OpenAI models at `/api/v2/llm/proxy/openai/v1` break tool calling in newer version of Code. This matters because the supply chain attacks on TanStack caused certificates to be reused which will eventually cause all users to have their malware detectors triggered and force an upgrade to a version where the certificate signatures match. Below is a full explination of the issue and how I fixed it with a LLM proxy that rewrites the calls locally before sending to AIP The issue was a **shape mismatch** between Codex, the Palantir OpenAI-compatible proxy, and MCP tool routing. Codex sent the MCP server to the model as a **namespace tool** : ```json { “type”: “namespace”, “name”: “mcp__codestrap_local_dev”, “tools”: [ { "type": "function", "name": "palantir_broker" } ] } ``` But the Palantir proxy/model response came back as a plain function call: ```json { “type”: “function_call”, “name”: “palantir_broker”, “arguments”: “{…}”
Ai-fde: ai chatbot studio
Hi, Are there any plans or timelines to expose AIP Chatbot Studio (Agent Studio) editing (viewing/modifying agent configurations, tools, and retrieval context) through AI-FDE, similar to how Pipeline Builder is supported today? Thanks! 2 posts - 2 participants Read full topic
Foundry workers for SQL Server with named instance
has anyone successfully set up a foundry worker sync (via agent proxy) with a sql server that uses instance name? in particular, how did you configure the Data Connection UI and egress policy (since the port is not available at configuration time but dynamically derived at runtime) 1 post - 1 participant Read full topic
AIP Program for brazilians
Hi! I’m trying to sign up using the link below, however, Brazil is not one of the listed countries. I’ve found some posts mentioning a “waitlist” but I wasn’t able to find any. AIP Signup 2 posts - 2 participants Read full topic
Transformation World 2026: The Best Data for Business AI
New Kyano® product innovations: Agentic AI for data migrations and unstructured data solutions Kyano Lorna supports customers with AI-powered project intelligence SNP and Palantir accelerate secure SAP transformations through strategic partnership High-profil…
Transformation World 2026: The Best Data for Business AI
New Kyano® product innovations: Agentic AI for data migrations and unstructured data solutions Kyano Lorna supports customers with AI-powered project...
Peter Thiel calls Pope Leo XIV a ‘Chinese communist agent’ over AI regulation push at Aspen Ideas Festival
Thiel's critique highlights the tension between ethical AI regulation and geopolitical competitiveness, underscoring global power dynamics. The post Peter Thiel calls Pope Leo XIV a ‘Chinese communist agent’ over AI regulation push at Aspen Ideas Festival app…
Peter Thiel Accuses Pope Leo of Serving as Chinese Communist Agent on AI
Billionaire tech investor Peter Thiel launched a series of controversial criticisms and predictions about AI and Western democracy during remarks at the Aspen Ideas Festival in Colorado on Tuesday. In his most provocative statement, Thiel accused Pope Leo XIV…