AI Strategy & Leadership
Seasoned AI and Analytics Leader with over 14 years of experience in AI strategy, business intelligence, and data science. Proven track record in leading AI-driven initiatives that optimize operations, enhance decision-making, and unlock new revenue streams. Adept at aligning AI strategies with business objectives and fostering cross-functional collaboration.
I wanted to share a bit about where the project is headed. The goal is to make this the go-to framework for any serious enterprise AI application where trust and depth of reasoning are non-negotiable.
Our immediate roadmap is focused on:
Expanded LLM Support: We're working on adding support for more open-source models (like the latest from Mistral and Llama) as well as improving integration with closed-source APIs for teams that use them in a hybrid environment.
More Data Connectors: We're prioritizing connectors for common enterprise knowledge sources like Confluence, SharePoint, and Salesforce to make ingestion seamless.
Performance Benchmarking: We're building a public benchmark to quantitatively show the performance lift of Graph RAG on complex, multi-hop questions compared to traditional vector-search RAG.
Simplified Deployment: We're finalizing a one-click Docker Compose setup and a Helm chart for easier deployment on Kubernetes.
We're building this in the open and would love to hear what the community thinks we should prioritize. What features would make this most useful for you?
I'm the founder of pmspace.ai. For over a decade, I worked in the architecture and construction world, and I was constantly frustrated by one thing: information chaos.
Every project generates a mountain of documents—hundreds of drawing sheets, 800-page specification books, endless reports, emails, and change orders. The critical data needed to make a decision was always somewhere, but finding it felt impossible. Using Ctrl+F across dozens of PDFs was a nightmare, and the ambiguity often led to expensive mistakes on the job site.
One night, after wasting an entire afternoon looking for a specific fire-resistance rating, I thought, "Why can't I just ask my project a question and get a direct answer with a source?"
That question led me to build pmspace.ai. I started with the feature I desperately needed myself: an intelligent document processing engine. You can upload all of your project's unstructured documents, and the system uses embeddings and LLMs to create a unified, queryable knowledge base. You can perform semantic searches ("find all walls with a 2-hour fire rating") or have a conversation with a specific document ("summarize the requirements for structural steel in section 05 12 00").
This document intelligence is the core of a larger platform we're building to be a true single source of truth for project teams. It connects directly to our workflow automation, BIM viewer, and other modules, all designed specifically for the AEC industry. Unlike generic AI wrappers, it's built to understand the complex, cross-referenced nature of construction documents.
We're a small, bootstrapped team. Our business model is a standard SaaS subscription, with pricing based on the number of users and data storage. There's a 14-day free trial with no credit card required so you can test it out properly.
This is still early, and there is so much more to build. I would be incredibly grateful for your feedback. I'd love for you to try uploading a complex document and see how it performs. What works? What's broken? What's the most frustrating information management problem you face in your own work?
https://bibinprathap.com/