Agentic Index: data analyst agents change log

Hex now lets Explorer users ask follow up questions inside published Generative Apps and continue investigating in an agent Thread.

Tue, 22 Sep 2026 12:00:00 GMT

Hex now lets Explorer users ask follow up questions inside published Generative Apps and continue investigating in an agent Thread. The same release adds CSV attachments to scheduled Generative App notifications. Exports use underlying project cell outputs, excluding transformations applied only in the app interface.

Why it matters: Business users can pursue new questions from an existing dashboard without rebuilding its context in a separate workflow. Teams distributing scheduled data should review the underlying outputs against recipient permissions, since visible app filters do not determine the exported dataset.

Impact: Medium. Verification: Verified. Type: Agent capability.

Evidence: Official changelog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Dataiku launched Cobuild on Snowflake, which combines Snowflake Cortex AI's model access with Dataiku's orchestration layer so users can…

Tue, 15 Sep 2026 12:00:00 GMT

Dataiku launched Cobuild on Snowflake, which combines Snowflake Cortex AI's model access with Dataiku's orchestration layer so users can build and validate AI workflows and agents in plain language.

Why it matters: Building agents where the data already lives avoids copying governed data into another tool. Snowflake customers should check how Dataiku's governance and Snowflake's access controls combine.

Impact: High. Verification: Partially Verified. Type: Integrations.

Evidence: News article

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Hex now lets teams publish and version evaluation suites that measure agent performance and test the effect of context changes.

Tue, 15 Sep 2026 12:00:00 GMT

Hex now lets teams publish and version evaluation suites that measure agent performance and test the effect of context changes. It also adds an option to skip agent approvals, and the chat agent can build Hex projects directly.

Why it matters: Versioned evaluations are how a data team proves an agent change did not make answers worse. The option to skip approvals should be set deliberately, workspace by workspace.

Impact: High. Verification: Verified. Type: Observability / auditability.

Evidence: Official changelog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Kenmei introduced Autonomous AI Agents designed to independently analyze telecom network data and autonomously resolve issues.

Sun, 13 Sep 2026 12:00:00 GMT

Kenmei introduced Autonomous AI Agents designed to independently analyze telecom network data and autonomously resolve issues. These purpose-built tools transition network operations from manual dashboard monitoring to autonomous problem-solving.

Why it matters: Telecom operators can reduce manual engineering tasks and accelerate issue resolution by deploying agents that act directly on network data.

Impact: High. Verification: Partially Verified. Type: Agent capability.

Evidence: Press release

Source: Agentic Index change log. Free to reuse with a link back. Terms.

OpenEvidence introduced a new family of medical AI models, including Osler, Sackett, and Snow, which offer varying levels of research…

Thu, 03 Sep 2026 12:00:00 GMT

OpenEvidence introduced a new family of medical AI models, including Osler, Sackett, and Snow, which offer varying levels of research depth and response speeds for clinicians. The company also integrated a precision oncology knowledge base into its specialized oncology AI sub-agent to provide genomic interpretations.

Why it matters: Healthcare providers can select specific AI models based on their immediate time constraints and research needs. The addition of the oncology knowledge base also allows oncologists to access specific genomic interpretations directly within their clinical decision workflows.

Impact: High. Verification: Partially Verified. Type: Agent capability.

Evidence: News article

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Hex released generative chart controls and shareable results for OAuth connections.

Thu, 27 Aug 2026 12:00:00 GMT

Hex released generative chart controls and shareable results for OAuth connections. Users can now leverage AI to generate and manipulate charts, while the OAuth update allows for the sharing of results derived from authenticated data sources.

Why it matters: Data teams can visualize insights faster using AI-driven chart generation. The ability to share OAuth-connected results also reduces friction in collaborating on secure data without compromising access controls.

Impact: Medium. Verification: Verified. Type: Agent capability.

Evidence: Official changelog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Lumi AI has released version 3.0 of its platform, introducing capabilities for handling complex, open-ended queries.

Wed, 26 Aug 2026 12:00:00 GMT

Lumi AI has released version 3.0 of its platform, introducing capabilities for handling complex, open-ended queries. The update enables the agent to perform multi-step analysis that users can edit directly within their workflows.

Why it matters: Enterprise data teams can now rely on the platform for more sophisticated analytical tasks that require sequential reasoning. The addition of editable analysis also gives users greater transparency and control over the insights generated by the agent.

Impact: High. Verification: Verified. Type: Agent capability.

Evidence: Official blog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

FinOpsly launched a new AI Cost Governance framework to help enterprises track and control AI expenditures.

Wed, 12 Aug 2026 12:00:00 GMT

FinOpsly launched a new AI Cost Governance framework to help enterprises track and control AI expenditures. The release includes Costix for pre-deployment cost modeling of AI workloads and Ask FI for attributing spending directly within Microsoft Teams and GitHub Copilot. It also introduces automated detection of cost spikes and runaway agents, routing alerts to accountable owners.

Why it matters: Enterprise buyers can now proactively model AI costs before deployment and attribute AI spending down to the user or application level. This prevents budget overruns from runaway agents and provides clear visibility for AI initiatives without requiring engineers to leave their workflows.

Impact: High. Verification: Partially Verified. Type: Agent capability.

Evidence: Press release

Source: Agentic Index change log. Free to reuse with a link back. Terms.

ketteQ announced the general availability of Quintus, an AI agent for supply chain orchestration that utilizes the PolymatiQ solver and…

Wed, 12 Aug 2026 12:00:00 GMT

ketteQ announced the general availability of Quintus, an AI agent for supply chain orchestration that utilizes the PolymatiQ solver and real-time Python code generation. The agent reasons through complex scenarios, executes unscripted tasks, and operates across existing enterprise resource planning platforms.

Why it matters: Operators can query live supply chain data in natural language to assess the impact of disruptions and retrieve immediate available-to-promise answers. Quintus deploys over existing systems like SAP IBP and Kinaxis, allowing organizations to adopt autonomous AI capabilities without replacing their current stack.

Impact: High. Verification: Verified. Type: Agent capability.

Evidence: Official announcement

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Corelayer announced its public launch as part of Y Combinator's Winter 2026 batch, formally introducing its AI-native production engineer…

Mon, 10 Aug 2026 12:00:00 GMT

Corelayer announced its public launch as part of Y Combinator's Winter 2026 batch, formally introducing its AI-native production engineer platform. The system monitors infrastructure and underlying data for silent anomalies, deploying agents to root-cause issues and propose fixes. This announcement marks the company's official debut for automating first-line production support in regulated environments.

Why it matters: For engineering and SRE teams in regulated sectors, backing by Y Combinator provides a strong signal of company viability and market traction. It offers institutional reassurance to buyers evaluating the platform for automating sensitive data pipeline monitoring and incident response.

Impact: Medium. Verification: Verified. Type: Agent capability.

Evidence: Official blog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

ThoughtSpot (Spotter): ThoughtSpot released version 26.8.0.cl, featuring a redesigned Admin Portal with an upgraded configuration UI and instant auto-save.

Wed, 05 Aug 2026 12:00:00 GMT

ThoughtSpot released version 26.8.0.cl, featuring a redesigned Admin Portal with an upgraded configuration UI and instant auto-save. The release introduces granular feature management, allowing Org Admins to self-serve feature toggles allocated by Cluster Admins.

Why it matters: Enterprise administrators gain significantly more control and efficiency when managing users, groups, and feature access. The shift to delegated hierarchical feature management is particularly valuable for large organizations with complex access requirements.

Impact: Medium. Verification: Verified. Type: Security / enterprise.

Evidence: Official documentation

Source: Agentic Index change log. Free to reuse with a link back. Terms.

ThoughtSpot (Spotter): ThoughtSpot officially launched AgentSpot, a new agentic workforce platform.

Tue, 04 Aug 2026 12:00:00 GMT

ThoughtSpot officially launched AgentSpot, a new agentic workforce platform. The platform enables users to build AI-driven workflows that make decisions, take actions, and deliver results across various business systems grounded in enterprise data.

Why it matters: This expands ThoughtSpot's capabilities beyond traditional BI and analytics into active workflow automation. Buyers can now leverage their data to trigger automated actions and decisions across their broader software ecosystem.

Impact: High. Verification: Verified. Type: Workflow orchestration.

Evidence: Official blog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

camelAI completely redesigned its coding agent architecture, migrating from per-user virtual machines to a serverless model using…

Mon, 27 Jul 2026 12:00:00 GMT

camelAI completely redesigned its coding agent architecture, migrating from per-user virtual machines to a serverless model using Cloudflare Durable Objects. The updated execution stack stores project files in SQLite and R2, replacing open-ended bash commands with sandboxed JavaScript Code Mode running in V8 isolates.

Why it matters: This infrastructure shift removes the fixed costs of always-on virtual machines, leading to more scalable and cost-effective deployments. Additionally, restricting open-ended shell access to explicit platform methods improves credential control and system security.

Impact: High. Verification: Verified. Type: Deployment / data residency.

Evidence: Official blog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

camelAI introduced a free tier powered by a self-hosted DeepSeek V4 Flash model running on AWS spot instances.

Tue, 21 Jul 2026 12:00:00 GMT

camelAI introduced a free tier powered by a self-hosted DeepSeek V4 Flash model running on AWS spot instances. The architecture routes requests through a Cloudflare AI Gateway with automatic failover to Azure's hosted DeepSeek service.

Why it matters: Evaluators can now test the platform's data analysis capabilities at no cost. The dedicated routing and failover infrastructure ensures that free-tier users experience consistent performance during their trials.

Impact: High. Verification: Verified. Type: Pricing / packaging.

Evidence: Official blog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Genspark AI Workspace: Genspark launched AI Workspace 6.0, introducing a persistent memory layer called SecondBrain that consolidates context across emails…

Mon, 20 Jul 2026 12:00:00 GMT

Genspark launched AI Workspace 6.0, introducing a persistent memory layer called SecondBrain that consolidates context across emails, documents, and meetings. The release also includes the Super Agent intelligence engine, a GenTeam collaboration layer for human-agent teamwork, and a companion hardware device called SecondBrain Note to capture offline conversations.

Why it matters: Enterprise buyers can leverage this unified memory architecture to reduce repetitive context-setting and allow AI agents to act on cross-platform data natively. The addition of hardware and dedicated team layers provides a structured environment for integrating autonomous agents into daily corporate operations.

Impact: High. Verification: Partially Verified. Type: Memory / state.

Evidence: Official blog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Actian expanded AI Analyst (formerly Wobby) with a Data Steward Agent that synchronizes business definitions and enforces data policies.

Sun, 19 Jul 2026 12:00:00 GMT

Actian expanded AI Analyst (formerly Wobby) with a Data Steward Agent that synchronizes business definitions and enforces data policies. The update also introduces Data Observability Agents to monitor pipeline health and a Model Context Protocol (MCP) server, allowing third-party AI agents to query governed business logic via open standards.

Why it matters: Data teams can now safely expose their semantic layer to external AI workflows without losing policy controls or context. This prevents external agents from operating on stale or ungoverned data, ensuring enterprise compliance and reducing hallucination risks.

Impact: High. Verification: Partially Verified. Type: Security / enterprise.

Evidence: Official announcement

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Genspark AI Workspace: Genspark redesigned its AI Slides canvas into a full presentation workspace, adding a left outline rail, tabbed project views, and…

Fri, 17 Jul 2026 12:00:00 GMT

Genspark redesigned its AI Slides canvas into a full presentation workspace, adding a left outline rail, tabbed project views, and comprehensive zoom controls. Users can now watch the AI generate slide decks page by page, verify figures, annotate changes like a printed draft, and generate speaker notes before using the Presenter View.

Why it matters: This interface overhaul gives users greater visibility and editorial control over AI-generated presentations. By enabling real-time human intervention and markup, teams can ensure the final output strictly adheres to their messaging and formatting standards before presenting.

Impact: Medium. Verification: Partially Verified. Type: Human approval / guardrails.

Evidence: Official changelog

Source: Agentic Index change log. Free to reuse with a link back. Terms.

Pluvo updated its Terms of Service to explicitly prohibit the use of customer data and model outputs for training or fine-tuning AI models.

Mon, 06 Jul 2026 12:00:00 GMT

Pluvo updated its Terms of Service to explicitly prohibit the use of customer data and model outputs for training or fine-tuning AI models. The revised terms clarify that customers retain full ownership of their data and the generated analysis. Pluvo also formalized its policies around the use of third-party AI sub-processors and cloud infrastructure.

Why it matters: Finance teams dealing with highly sensitive company data require strict privacy guarantees before deploying agentic workflows. By contractually ensuring that data will not bleed into public or aggregate foundation models, the vendor removes a major legal blocker for enterprise adoption.

Impact: High. Verification: Verified. Type: Security / enterprise.

Evidence: Official documentation

Source: Agentic Index change log. Free to reuse with a link back. Terms.