Promptius AI
Build production-grade AI agents using natural language
143 followers
Build production-grade AI agents using natural language
143 followers
Promptius is an AI-powered development environment designed to build autonomous agents based on user prompts.
Interactive
Free Options
Launch Team / Built With
ArticosTest your messaging on your audience, right inside Claude or ChatGPT
Promoted
Hey Product Hunt 👋 Kanishk here. Last year I fell deep into AI agents. One late night, after an 8-hour coding marathon building a multi-agent system, I thought: What if an agent could build agents? That idea became Promptius.
Promptius turns natural language descriptions of business processes into production-grade AI agents — in minutes, not months. Non-technical users can “vibe-code” an agent through conversation. Technical folks get transparent langgraph python code, full control in a built-in IDE, and one-click deployment.
We handle the hard stuff — architecture generation, prompt & context engineering, 10,000+ tool integrations, and built-in testing — so you can go from idea to reliable, context-aware, multi-step agents fast. Agentic engineering is no longer just for AI experts.
@kanishk_gupta_2000 Congrats on the launch, love the production‑grade angle. How do your agents build multi‑tool workflows with reliable observability / failure recovery, and can teams define reusable guardrails/policies that prompts can’t override?
@tonyabracadabra thanks for the support! to answer your question: Promptius focuses on building Langgraph agents, and Langgraph comes with out of the box support of Langsmith, which is one of the best tools out there for observability, we can vouch for it after months of testing our own agent and making improvements and tracking performance via langsmith.
Regarding failure recovery and guardrails, our team is working on enhancing the process so that the furnished agents have these batteries built in. We are happy to help users to guide their projects to production in the meanwhile. Our generated prototypes can also be edited within the built in IDE and users have complete control over their agents!
Everyone I am sorry for replying late, looks like a product hunt glitch because many of the comments are coming up to my feed after 3-4 hours!
I wanted to try this but there seems to be a bug. Getting an 'Internal Server Error'.
I tried on a different network too. Kindly look into it.
@ashok_nayak Hey Ashok, apologies for delay in the response, I am not sure how I missed your comment! Yes we were not prepared for the surge in traffic xD We have now scaled up the server and constantly monitoring the service. Please do try it out! Thanks for informing us <3
Seems like a valuable problem...lots of talented teams are building tools in this space. What are the others getting wrong? E.g. how is Promptius better than, say, @Okibi ?
@rajiv_ayyangar Definitely there are so many teams out there working on this problem and we are happy to see it grow! We think Promptius differentiates in a few key ways from other agent builder plays:
We are not investing in building our own sdk, we are banking on langchain/langgraph which already comes with a robust open source community.
This frees up to work as a meta-programming platform, and focus more on agentic engineering and parts of it such as workflow design and architectures.
Our tooling support is extensive.
We have also introduced a novel agent reinforcement learning and simulation sandbox, all our agents get tested in a sandbox based on user's requirements, and tested till they get optimized to the user's requirement specification!
Great job! I like how it quickly understands which features I need and which tools to use. Will explore it more.
I've also tried AHITL, a few points there:
- I wish when I fill in the questions in the chat box with AI my answers stay in the chat instead of disappearing
- When the model asks me to fill in the details, I'd like to have an option to continue chatting - now I'm stuck if I have to fill in the details
@steffanb Hey Steffan I really appreciate you engaging with Promptius as well as AHITL! We are super excited about AHITL and would love to grow it as a standard for adoption.
Your points make total sense, we have to make improvements over our UI/UX, thank you for your feedback! <3
Sounds very interesting!
I spent most of yesterday playing with OpenAI agent kit, will be interesting to see how this stacks up.
In any case, awesome job! Congrats on the launch!
@ivan_sem Yes the space is developing rapidly and I think we are taking a different approach than visual workflow builders. We believe that given how AI excels at writing code, allowing users to go home with code has a separate value add. This will eventually help us move faster and provide even more value to our users!
All the best guys! I really love the dry run and scenario trigger design mode; it really helps you put things into perspective and scope out your implementation in ways you would have thought of while deep into implementation. Kudos for the time save! Looking forward to even better integrations and above all, the pay per use in built hosting (as long as you keep it decently priced XD)
@ansh_riyal1 Hey Ansh, we built Promptius keeping human in the loop at heart, gaining enough information from users beforehand to be able to build solid agents! Happy to see this resonates with you. We currently support an agent simulation sandbox and we are working on our deployment capabilities, with focus on reasonable hosting strategies depending on what are the availability requirements of our users.
Good luck guys! Does it support multi-agent collaboration out of the box?
@roozbehfirouz hey, yes it does! What promptius does best is separation of responsibilities, separating different units of works to different workers/agents. This is driven by our workflow design generation process.