Who we are
Vipul Mehra, founder. Vipul builds the systems. He has spent more than 10 years in data and AI across fintech, healthcare, e-commerce and gaming. He built and runs a call-scoring system that handles 1,000+ calls a day for a US financial services company.
Priyanka, co-founder. Priyanka is an AI researcher with a software engineering background. She runs model evaluation, which means every model is tested on real examples before it ships.
What we build
| Service | What it does |
|---|---|
| AI Sales Coach | Every sales call scored 1–10 on 8 measures. A coaching email to each rep every weekday morning. Compliance issues flagged on a dashboard, most the same day. |
| AI Assistant | A private assistant that answers only from your company's own documents. |
| AI Analytics Agent | On request. |
| AI Receptionist | Early access, on request. |
How we work
- One small step first. Every service starts with one small, fixed-price step: a $750 Blind Match for call scoring, or a $750 Hundred Questions for an assistant. You see results on your own data before committing to anything bigger. See steps and prices.
- You own the accounts. Your AI running costs (OpenRouter, AWS) are billed to your own accounts, typically cents per call. We never mark up your AI bill. If you'd rather we host it, AI usage is passed on at cost.
- We reply within 1 US business day.
- We say what we have measured, and what we haven't. Each case study has a "What we did not measure" section. Start with the call-scoring case study. There is also an AI tutor case study.
- Open-source models first. They are cheaper, you can inspect how they work, and you are not locked to one vendor.
- We write down what we learn. We write it up in our guides. We also publish our own AI visibility score on the log.
Contact
Email info@northstrdata.com or book a 20-minute call. We reply within 1 US business day.