About us
We got tired of forecasts nobody trusted.
Every revenue team we talked to had the same problem. They had data. They had dashboards. They had gut instinct. What they didn't have was a forecast their whole team believed in — one that was accurate enough to make real decisions from, and transparent enough that the CFO didn't just wave it away.
So we built LTprophecy. Not a dashboard. A reasoning engine for revenue — one that combines ensemble machine learning, autonomous AI research, and causal knowledge graphs into a single platform that gives B2B teams an actual intelligence advantage.
CogniRegal Lab (OPC) Pvt. Ltd.
We're a product-first company incorporated in Mumbai, Maharashtra under the Companies Act 2013. Our name comes from “cognition” — understanding — and “regal” — setting a higher standard. We believe that understanding your revenue deeply, and acting on that understanding with confidence, is the highest-leverage thing a revenue team can do.
We're a small, focused team. We ship fast, we listen closely, and we take the reliability of your revenue data more seriously than almost anything else.
Founded
2025
Location
Mumbai, Maharashtra, India
Entity
OPC Private Limited
CIN
U62020MH2025OPC462835
What we've built
Ensemble ML Forecasting
XGBoost, LightGBM, and BayesianRidge stacked and auto-tuned. Every model is backtested on your actual data with confidence intervals that are honest about uncertainty.
AI Research Agents
LangGraph-powered agents that crawl financial filings, market signals, and competitor intelligence — then propose updates to your knowledge graph.
Causal Knowledge Graph
Model relationships between business factors in a Neo4j graph. Run Bayesian simulations. Understand which levers actually drive revenue before making decisions.
Pipeline & Data Tooling
Visual DAG editor powered by Dagster. Feast feature store. MLflow experiment tracking. Everything a revenue team needs to operate at ML scale.
How we think
What we believe about forecasting.
Honesty over comfort
We'll show you confidence intervals that are wide when the model is uncertain. We won't narrow them to look impressive. A forecast that's honest about its limitations is more useful than one that overpromises.
Explainability is non-negotiable
Every prediction LTprophecy makes can be traced back to specific features, weights, and causal relationships. We believe you should be able to push back on our model — and we make that possible.
Revenue teams are not just spreadsheet operators
CFOs, VP Revenue, and analysts deserve tools that work as fast as they think. We build for speed of insight, not just depth of data.
Infrastructure first
Security and reliability are not features — they're the product. We don't ship new capabilities without earning the infrastructure trust first.
We're building something hard.
Combining ML forecasting, autonomous agents, and causal knowledge graphs into a product that actually works requires people who love the details. If you're a machine learning engineer, product designer, or full-stack developer who wants to work on something with real commercial stakes — we want to hear from you.