Verified purchaser
Great concept, but way too many bugs and stalled development
I honestly don't understand why this product has so many glowing reviews when the platform is currently riddled with bugs. I started out really excited about TeamPal because the concept is amazing, but after putting the workflows and agents to the test for over a month and a half, it's just not ready for serious work.
Here are the main dealbreakers for me:
1. Stalled Development: The last major update was back in January. It’s been several months now without any noticeable improvements, new features, or updated LLM models. It feels like development has completely ground to a halt.
2. Unreliable AI Iterations: When running agents and switching between models, the system constantly throws errors saying it can't process the request. Instead of saving me time, I end up wasting hours troubleshooting and repeating tasks.
3. Broken Knowledge Base Integration: This was supposed to be the tool's biggest strength, but it doesn't work intentionally. You can upload all the documentation you want, but if you don't manually point the agent to the exact files every single time, it completely forgets them and acts like a generic AI. Digging through 5 to 10 documents to guide it manually defeats the entire purpose of automating the workflow.
The potential is definitely there, and they might have a good niche in the future, but right now it's too frustrating for a long-term commitment. With no bug fixes or real support updates after weeks of testing, I decided to request a refund before my window closed.
Jeff_Pham
Jun 1, 2026Thank you for taking the time to share such detailed feedback, and we're sorry your experience didn't meet expectations. Honest reviews like this genuinely help us improve, so we appreciate it.
A few things we'd like to address:
On development: we understand it may have looked quiet, but the team is actively working behind the scenes, and updates including new LLM models and improvements are on the way. We'll be much more visible about our roadmap going forward.
On the model-switching errors: this is a known issue we're actively fixing. In the meantime, switching to a stable model (such as Claude) usually avoids the errors, and we're happy to help you set that up.
On the Knowledge Base: thank you for flagging this, the behavior you described isn't how it's meant to work, and improving retrieval so agents reliably use uploaded docs without manual pointing is a current priority.
We completely respect your decision to request a refund. If you're ever open to giving it another try, we'd love the chance to show you the improvements, and we'd be glad to support you directly. Thanks again for the thoughtful feedback