# Article Name Torii AI for SaaS- Eko Agent, GPT Actions & Parsing # Article Summary Tori's AI SaaS solution turns contracts into data and uses Eko, MCP and GPT Actions to automate workflows, cut spend and risk # Original HTML URL on Toriihq.com https://www.toriihq.com/articles/torii-ai-features # Details How is Torii parsing thousands of messy SaaS contracts and delivering 90-day optimization plans through five core Eko interactions? Discover AI contract parsing that extracts renewal dates, license tiers, clauses, and true cost breakdowns in seconds. Uncover instant answers to cost questions for apps like Algolia, plus code-backed insights to rank the ten most expensive apps and eliminate surprise renewals. In this video, Ori Amarcio, co-founder and CEO of Tori, shares how Eko, MCP, and GPT Actions turn SaaS data into instant answers and automation. After quick poll takeaways on preferred LLMs and prompting style, see contract parsing feed clean data into Eko’s five interactions for smart support, quick answers, advanced insights, training, and real savings, then watch MCP bring answers into external chat while GPT Actions check breach history, compose emails, and streamline workflows. A must-watch for IT, procurement, security, and FinOps teams that need to cut waste, prevent renewal surprises, and act on trusted data before 2025 renewals land. This article was originally a video (YouTube link here [https://www.youtube.com/watch?v=7Rd37zk9Pn4]). Below is the full transcript: I'm Uri Haramati co-founder and CEO of Tori, and I'm passionate about AI. What we're sharing today is not just another product update — it's a major step forward in SaaS management. In 2025, the year of agents, AI is moving from buzzword to game changer, and Torii is building that future around a powerful loop of data, intelligence, and autonomous actions. We collect rich SaaS data across environments, supercharge it with AI that understands context and intent, and empower you to act instantly and safely. This is working now for customers and forms the foundation of the features we'll demo today. First up is AI contract parsing. Launched in late 2023, this capability laid the groundwork for everything that followed. We parsed thousands of contracts across our customer base, which improved accuracy and helped the model learn from real-world use cases. SaaS contracts are often messy and buried in PDFs. With AI contract parsing, we automatically extract critical details such as renewal dates, license tiers, clauses, and true cost breakdowns. Static files become dynamic, actionable data accessible throughout the Torii platform, eliminating manual entry and preventing surprises at renewal time. In the demo, a contract was uploaded via drag-and-drop. The parser identified contract dates, total yearly cost fields, cancellation notice periods, authors, and line-item details. Once saved, that contract data is available for automation, reporting, and other workflows across Tori. Next is Eko, our agentic SaaS co-pilot. Eko is a full-fledged AI agent powered by a multi-agent architecture: a network of specialized agents that interpret questions, pull data, generate insights, guide decisions, and create code when needed. Together they produce a unified experience. Eko supports five core interaction types. The first is smart support: if you ask how to integrate a tool with Tori, Eko provides step-by-step instructions, required permissions, and prerequisites instantly. This eliminates digging through documentation. The second is quick answers. Eko can query your environment directly — for example, "How much are we paying for Algolia?" — and return a precise result by looking at contracts, invoices, and usage data. It can also surface follow-up actions for deeper analysis or optimization. The third capability is advanced insights. Eko can connect the dots across usage and spend and even generate code to produce complex analyses. For example, it can compute your ten most expensive apps and present a comprehensive view of cost drivers and opportunities to act. The fourth is a training and best-practices agent. Ask Eko what workflows you can create in Torii and it will recommend workflows like license optimization, contract renewal management, and underutilization detection. It acts as an onboarding and enablement assistant, offering pro tips and step-by-step guidance. The fifth is deep-learning collaboration. Ask a broad question such as "Where can we save money on our SaaS apps?" and Eko will scan your stack, compare usage and spend, apply Torii best practices, and recommend specific savings opportunities. Eko can write code snippets to analyze data, produce a 90-day optimization plan for a given app suite like Google Workspace, identify stakeholders to involve, and recommend automated workflows to reclaim licenses. A quick note: these advanced features depend on accurate data in Tori. Foundational capabilities like contract parsing and automated data ingestion make Eko’s insights possible. We focus heavily on reducing friction for getting that data in. We also released an MCP integration (Multicontext Protocol). MCP lets external chat environments or tools talk to Torii so you can query Torii from other platforms. For example, you can ask in your external chat, "How much are we paying for Salesforce?" and Torii will return contract cost, the active contract, and the owner without leaving the chat. The final major feature is GPT Action, which lets you call an LLM directly within a workflow. Instead of relying solely on static checks, a workflow can query an LLM for external context, perform analysis, or generate human-facing content. One example: when a new app is discovered, a GPT Action can ask whether any security breaches have been reported for that app, summarize the nature of a breach in plain language, and update application fields with the LLM's response. You can then surface that information in a security compliance dashboard widget. GPT Actions can also compose emails or notifications. For instance, if a breach occurred in the last two years, the workflow can auto-compose an email to the security team with a summary and remediation steps and then send it or queue it for review. This capability unlocks many automation scenarios customers have implemented, including extracting company size, sentiment from support portals, and other nonstandard signals. Those are the major generative AI features we’ve released so far: AI contract parsing, Eko, MCP integrations, and GPT Action. We have other smaller enhancements in the product and a roadmap that includes an AI workflow builder to create entire workflows from natural language, predictive renewals and risk scoring, creating widgets and dashboards from prompts, and richer Eko capabilities that increasingly let you work with Torii via natural language. In the Q&A, we addressed how Eko differs from other AI chat tools. Eko combines knowledge and the full context of your SaaS ecosystem with Tori’s best practices and robust data. That combination, plus speed, is what makes Eko unique. For teams early in AI adoption, start with discovery and connect your apps to get a clearer inventory. From there, use insights and build simple workflows. Alternatively, start by asking Eko what you should connect next; Eko can guide discovery and recommend priorities. We will send a recording of this webinar after it is rendered and uploaded, and we will contact the raffle winner following the session. If you have additional questions, please reach out to your CSM; they can help you enable these features and get them running in your environment. Thank you for joining us today. We appreciate your time and interest in Tori’s AI capabilities.