Deep Talk Reviews

i love this. wondering when they will come back so i can upgrade to tier3!

if you no buy this, you are missing out! this ai is very very good.

i buy tier 1. hope they come back to appsumo soon.

I know this tool is going to save me hours on work I usually do manually.

Quantitative data is valuable and this tool won't replace analyzing the data, but it will help calculate frequency of keywords in the data and help me visualize the results.

Oh... and the need for customer support was because I (horror of horrors) misspelled my email address and there was no easy way to change the account setup yet. They had it fixed within minutes.

I'm an online course developer, coach, and content creator.

I can imagine most business people have a multitude of channels they use to stay in touch with their audience/customers, for example, there's a wealth of valuable text data in email, live chat, surveys, quizzes, social media channels, etc. All of these channels contain a wealth of information that you can gain important insights to help grow your business. I've always wondered how I could make sense of all that data.

The problem is that all of these channels contain unstructured data. Unstructured data cannot be displayed in rows, columns, and relational databases. It's complex to make any sense of this data because it's difficult to process them in comparison to structured data.

Deep Talk has been a wonderful aid in helping me sift through my unstructured data. And all without needing to write a single line of code.

So far I tested the app with simple datasets, but I can already feel the underlying power of the app, plus it has multiple use-cases.

Pros
- Incredibly user-friendly platform that takes you by the hand to configure models without writing any code.
- Classify, categorize and detect relevant topics
- The data is presented in 3 levels of depth

Cons
- Sentiment analysis would be a wonderful addition

Overall, Deep Talk's UI/UE is well thought out, The limits are good, and the deep insights are valuable. The platform has multiple use-cases. Definite recommendation!

Thank to the Deep Talk Team and Appsumo for the bring this deal!

So most of you are using this for marketing and support. That is a great use case for it. I am using it for financial data. I just uploaded the last year worth of FOMC statements to see if Deep Talk matched up with my models and it does, but it does so much more.

The charting on this is fantastic. If you cant figure out what is and is not important in about 3 minutes of this tool then you are blind. It highlights what is actually being said, and the more data you can feed it the better.

I only did like 8 rows of data-but each row was actually a 5-8 paragraph statement. I can't wait to feed it the last 30 years of statements and then start feeding it not just earnings calls but also news articles, and statements from CEO's.

The limits are good, the interface is very clean, and the insights are priceless...or at least they can be if you act on them.

Initially was drove me to this tool was the possibility to train data models on our Intercom chats and tickets to see topics, issues, and trends visualization.
While we have a good idea of what those are, the insights were pretty interesting and allowed me to build SOPs to list down the most recurring customer questions and issues.

Initially, I had a few issues training the model on our data, and Sergio, Juan, and the Deep Talk team helped me to improve the model in a prompt manner. I really appreciate all the conversation and help from them.

For me, Deep Talk can find a great use for multiple cases:

- Customer Support: see the most talked topics to tap into bottlenecks, recurring issues your customer may face and improve the support and the product.

- Content research: scrape social media posts and comments and analyze the most talked about topics and come up with fresh content ideas.

- Competitive intelligence and market research: what their customer say about them and the most recurring pain point or feature requests

- Opinion polls, journalism: dissect trends, opinions, and write about them.

At the moment it's still a beginning product, I hope more visualization and training models will be added to discover even more insights from the data.

Limits are pretty generous as most said here. Plenty of use cases and ideas to try with this!

All in all, I can recommend this product and the team behind have a great attitude and proactively answered all my question.

All the best for your launch, very happy to support you guys!

I've been testing out DeepTalk - it's good, generally fast and easy to use. This is an excellent platform and the monthly renewing quotas are super generous. This is a really intriguing tool for copywriters and I am using it in tandem with GummySearch. If customer quotes, sentiment analysis, keyword frequency, relevancy and data visualization are your jam, I’d hop on this.

DeepTalk are very open to suggestions and their chat support is phenomenal, quick, friendly and helpful. They are working on video modules to help onboarding soon. On my first attempt, I needed (a lot of) chat support uploading a CSV file. Support was on it. Very impressed.

Overall, Deeptalk’s UI is streamlined. One area def needs improvement: there is currently no option to delete a project once created, but you can easily edit a project’s name and then upload a new training model/ csv data. Still, I like deleting unnecessary projects to keep my dashboard neat.

There are some quirks to uploading a CSV file - the platform includes demo files and dialogue boxes to guide you - super helpful. It's key to follow DeepTalk's CSV formatting instructions so you can upload and then map a compatible CSV into their system. This took me a few tries, but I got the hang of it quickly. Now I’m a pro lol.

If you are uploading simple TEXT data, such as product reviews, you only need 2 columns in your CSV file.

If you are uploading CONVERSATION data, such as chat text, you need to have your data in 4 very specific columns (Conversation ID, Timestamp, Content, and Person Type. For the 'person type' column, create both a 'customer' and 'agent' in your CSV. If no agent 'person type' exists in your data, create one (I just created a line item in my CSV and made myself an agent). Then DeepTalk's system can parse the data correctly. They also have native Intercom integration and more integrations are on their roadmap. Again, if you select the text data option above, you can skip all the customer/ agent weirdyness.

DeepTalk does exactly what it promises - and it’s fun to use. I’m really glad I found this one and I’m excited to see how it develops.

Based on a quick session; have found the service reliable and very easy to work with for CSV imports, certain areas require additional work and their fantastic support has already taken notes onboard from the list below which I hope most can be included or improved in the upcoming updates:

1. Clicking outside certain boxes by error loses the work, have to start again, please resolve this, perhaps warn before closing
2. Reports by email would be good to the email address of choice for each
3. Direct Infinity (StartInfinity) integration
4. Need to be able to get access different teams with different roles/permissions (view only, manage and etc.) accessing different Workspaces
5. More assurance on it being GDPR compliant
6. On hover effects/interactions from the data visualisation (perhaps use www.amcharts.com which is JavaScript Charts & Maps or something similar for graphs etc.)
7. Report to PDF or Interactive file such as PowerPoint
8. On Trends when clicked on the back arrow/delete, unable to get it back
9. Ability to use URL for live data feed (e.g. from Amazon reviews, Tweets, Facebook, Google Reviews and etc…)

Having got Hazlo and Gyana earlier and struggled with them, I was not sure about this. Happy to say, at least in terms of ease of use, Deep Talk is superior. Setup is fairly straightforward and once you are logged in there are the onboarding panels which, for me, was sufficient to get started on my first project. Now while the journey is fairly easy some of the language could be improved to make it more intuitive. For example, after uploading my data set I was left confused as it didn’t allow me to proceed. Only when I pressed the ‘upload new file’ did it upload. I thought it should have been ‘confirm upload’ if that was the purpose. There are such minor improvements that can be solved easily.

Anyway once the data was uploaded it took about 3-4 minutes during which it just showed ‘training’. Again here I wasn’t sure what was expected of me until it was finished. But once it was done, the outputs seemed impressive. The sunburst chart sure is attention grabbing. It also lets you know each time how much rows/credits gets used up before processing, which I found helpful.

It’s kind of hard to assess the authenticity of the results with a few trials. I imagine it would take at least a few dozen data sets and training during which the AI gets better before one can truly assess it.
Having said that, I must say that the use cases of these go beyond just customer support. Sure the integration is with Intercom but you can upload your own csv data. Imagine what that can yield. I have a few sentiment analysis tools and currently looking to see if I can get common data sets that I can feed to both and if it succeeds then I can use it for listening, sentiment analysis, SM growth, maybe even drive sales. Another idea that struck me is using Deep Talk in tandem with Market Muse for competitor analysis!!

All the best to the team. I hope you guys take the feedback from this community to shape your roadmap- it will be a win-win ✨