# Searching Etimad Tenders with AI: A Full Comparison Against Manual Search ## Introduction Everyone working in Saudi government tenders knows the daily routine: open the **Etimad** platform, filter results by city and activity, review dozens of titles, then repeat the process the next day looking for any new tender. With AI assistants like **Claude**, **ChatGPT**, **Perplexity**, and **Grok** entering this space via the **MCP protocol**, there's now another path: ask a question in natural language and let the assistant search, filter, and sort for you. This article compares the two approaches in detail, and clarifies when AI-powered search is better, and when manual search remains necessary. ## What Does Manual Search on Etimad Look Like? Traditional search on the Etimad platform follows fixed steps: - Open the available tenders page. - Select a city or region from a dropdown list. - Select an activity or sub-activity. - Specify the tender type (open, limited, small/medium enterprise tender...). - Review results page by page, opening each tender to confirm its details. This method is precise and reliable, but it consumes time that multiplies with every repeated search, especially when you're searching across more than one activity or city. ## What Does AI-Powered Search Look Like? Once you connect any AI assistant to TendersAlerts' MCP server via: `https://api.tendersalerts.com/mcp `it's enough to type a question like: > "Search for open IT tenders in Riyadh, sorted by nearest submission deadline." The assistant automatically calls the **search-tenders** tool and returns sorted, ready-to-use results, with the ability to request the next page or adjust the filters with another sentence, such as "also add the telecom activity" or "exclude limited tenders." ## Comparison Table: Manual Search vs. AI-Powered Search CriterionManual Search on EtimadSearch via Claude/ChatGPT/Grok/Perplexity**Time required**Minutes to tens of minutes depending on the number of filtersSeconds per query**Query format**Fixed dropdown listsNatural language, Arabic or English**Adjusting the search**Re-filtering from scratchOne follow-up sentence in the same conversation**Combining search with analysis**Not directly availableCan move straight to market or competitor analysis in the same chat**Login requirement**Etimad accountBasic search works without login via MCP**Accuracy of reference results**Always highAlso high, since data is read live from the same source## Which Assistant Should You Choose? Claude, ChatGPT, Perplexity, or Grok? The good thing about the MCP protocol is that it's a unified standard: one TendersAlerts server works with any assistant that supports **Streamable HTTP**, which means the choice depends on your personal preference for the assistant itself, not on the data it can reach: - **Claude**: suits users who need precise sequential steps, like searching, then reading documents, then analyzing, all in one long conversation. - **ChatGPT**: good for those who already use it for other daily tasks and want to add Etimad data as one more tool among them. - **Perplexity**: its users tend to combine Etimad search with general web search in one conversation, such as comparing a tender against information about the owning agency from other sources. - **Grok**: a good option for those already working within the X ecosystem or who prefer its interface. The connection method is similar across all four: paste the server address and complete login via OAuth 2.1. See the [full connection guide](/docs/mcp) for exact details for each assistant. ## A Practical Scenario: A Workday Before and After AI Imagine a tender officer tracking opportunities across three different activities (IT, maintenance, construction) in five cities. Before connecting to MCP, their typical day started by opening Etimad and repeating the filtering process fifteen times (three activities × five cities), then manually reviewing the results to exclude duplicates or irrelevant ones. This process alone could take half an hour to a full hour every morning. After connecting, they write a single question: **"Search for open IT, maintenance, and construction tenders in Riyadh, Jeddah, Dammam, Makkah, and Madinah, sorted by nearest deadline."** They get a unified list in seconds, and can immediately follow up with "exclude anything due more than two weeks out" to narrow the results further. This shift doesn't just save time — it also reduces the risk of missing a tender because of a forgotten filter in a particular city. ## When Does Manual Search on Etimad Remain Necessary? AI-powered search doesn't eliminate the need for the official platform entirely. Some tasks remain better suited to direct execution on Etimad: - **Submitting the bid itself** and uploading official documents, a legal procedure that must happen directly on the official platform. - **Final confirmation of classification and qualification** before bidding on a tender that requires specific certificates. - **Reviewing any urgent update** announced directly on the platform before it gets indexed anywhere else. AI via MCP is a tool for accelerating search, analysis, and understanding — not a substitute for the official submission procedures. ## Where Is Search in Government Tenders Heading? The global market trend is clear: generative AI assistants have become the first entry point for many users, ahead of any traditional search engine or specialized platform. For the Saudi government tenders sector, this means companies that master using these assistants connected to accurate data via MCP will be able to spot opportunities and make decisions faster than companies still relying entirely on manual browsing. This doesn't mean manual search disappears — it turns into a verification and exception tool, while AI handles most of the daily routine screening. Companies building this new habit now will find themselves in a much better position once relying on AI assistants becomes a common standard across the entire sector in the coming years. ## Advanced Search Questions to Try Directly - "Search for maintenance tenders in Jeddah and Dammam together, sorted by most recently published." - "Give me open tenders for the Ministry of Education only, in the IT activity." - "Search for small and medium enterprise tenders in the construction activity in the Asir region." - "Convert the city name Abha into its ID first, then search it for open tenders." (calls **taxonomy-lookup** then **search-tenders**) ## Building a Daily Search Habit Without Effort The best way to benefit from this shift isn't a one-time use, but making it part of your regular morning routine. Since the connection between the assistant and the MCP server remains active after the initial setup, you can simply open a new chat every morning and repeat the same search question with a small date adjustment, instead of going back to browse dropdown lists again. Over time, you'll learn how to phrase your questions more precisely to get results closer to what you need on the first try. ## Frequently Asked Questions **Does AI-powered search require login?** No, the **search-tenders** and **taxonomy-lookup** tools work for free without login in any assistant connected to the server. The other tools, like market and competitor analysis, require an [active subscription](/plans). **Do results differ between the four assistants?** No, all of them call the same MCP server and get the same data from the source; the only difference is how each assistant presents and explains the result. **Can you search in English?** Yes, via the **locale** parameter, which the assistant sets automatically based on your question's language, but note that some fields, like alert keywords, remain tied to the tender's Arabic text. ## Conclusion Searching Etimad tenders with AI via MCP doesn't replace the Etimad platform — it adds a fast layer of filtering and understanding on top of it. Instead of repetitive dropdown lists, you ask in natural language and get a sorted result in seconds, with the ability to move straight into market or competitor analysis in the same conversation. [Try searching Etimad tenders through your favorite AI assistant now](/mcp).