Tavily Enterprise
TavilyReviews from AWS customer
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
1 review
from
External reviews are not included in the AWS star rating for the product.
Tavily made our company enrichment pipeline actually accurate
What do you like best about the product?
Three things stand out for my use case:
(1) Concise and relevant in the same response. I run search across hundreds of thousands of companies, including a long tail of small businesses that Google just doesn't surface well. Tavily gives me results that are filtered down to what's actually about the right company, without burying me in noise or doing so much post-processing that I lose control of the raw signal.
(2) The relevance score on every result. I use it to filter to a clean threshold so my downstream LLM gets exactly the right amount of context, not too little and not too much. This was the missing piece in every other solution I tried.
(3) Integration was effectively a URL swap. I had a SERP-based pipeline already running. Replacing it with Tavily was a five-minute change. I deployed it, it worked, and I haven't had to touch the integration since. Same story when I added Tavily to a LangChain agent we built later.
(1) Concise and relevant in the same response. I run search across hundreds of thousands of companies, including a long tail of small businesses that Google just doesn't surface well. Tavily gives me results that are filtered down to what's actually about the right company, without burying me in noise or doing so much post-processing that I lose control of the raw signal.
(2) The relevance score on every result. I use it to filter to a clean threshold so my downstream LLM gets exactly the right amount of context, not too little and not too much. This was the missing piece in every other solution I tried.
(3) Integration was effectively a URL swap. I had a SERP-based pipeline already running. Replacing it with Tavily was a five-minute change. I deployed it, it worked, and I haven't had to touch the integration since. Same story when I added Tavily to a LangChain agent we built later.
What do you dislike about the product?
Honestly nothing that's blocked me. The one piece of feedback is that I discovered some of the newer endpoints (like Crawl) only because I was preparing for a customer call. More visibility into what's launched recently and good worked-examples for those endpoints would help me get to the next use case faster.
What problems is the product solving and how is that benefiting you?
We pull structured information about hundreds of thousands of companies and feed it into the AI features our customers' sales teams use every day. Vertical, location, similarity, and other signals all start from this base layer. If that data is wrong, everything downstream is wrong.
My V0 used a traditional SERP API plus an LLM. Accuracy was OK but capped. I had a regression set of ten lookups it kept getting wrong, and every prompt fix in one place broke a different one. Typical hit rate was around two out of ten.
When I ran the same set through Tavily, all ten came back correct. That was a step change, not an increment, and it unlocked features we couldn't ship before with confidence.
Today Tavily powers our company enrichment pipeline and the search layer behind our Slack agent, which is used by sales reps at top B2B SaaS companies. The base layer is finally strong enough to build on.
My V0 used a traditional SERP API plus an LLM. Accuracy was OK but capped. I had a regression set of ten lookups it kept getting wrong, and every prompt fix in one place broke a different one. Typical hit rate was around two out of ten.
When I ran the same set through Tavily, all ten came back correct. That was a step change, not an increment, and it unlocked features we couldn't ship before with confidence.
Today Tavily powers our company enrichment pipeline and the search layer behind our Slack agent, which is used by sales reps at top B2B SaaS companies. The base layer is finally strong enough to build on.
showing 1 - 1