Launched this week
KnowYourCompany.ai
Uncover Hidden Insights, Invest with Confidence
99 followers
Uncover Hidden Insights, Invest with Confidence
99 followers
Stop investing on vibes ! KnowYourCompany.ai is the AI-native equity research platform that gives you the discipline the funds use, in plain English. 1. Find: is this worth my money? 2. Understand: what am I actually buying, distilled in one morning. 3. Stay safe: alerts the day something material lands. Deploy agentic AI to expand your coverage, offload manual review, and surface signals faster. Every fact linked back to the filing.
Hey Hunters 👋 Shams here, one of the makers.
Honestly, this started as a problem of our own. We were investing on the side, doing our own research, and tracking our own ideas. Then life got busy, and keeping up with everything we'd found just fell apart. The signal was there; we didn't have the time to stay on top of it.
So we started building agents to do it for us: surface ideas, dig into them, and follow up before the window closed. Coming from cybersecurity, it felt familiar. Pull scattered signals into one place and act before the moment passes.
That side project became KnowYourCompany.ai: a cognition platform to discover, research, and act on any of India's listed companies, with every claim cited to the exact filing, page, and line, so you never have to just trust the AI.
The lesson for us was that research isn't a time problem; it's a systems problem. You can't read 500 filings a day, but your stack can. It should be watching, filtering, and digging for you, so your time goes into thinking and deciding, not searching and verifying.
So I'm curious: what does your research stack look like today? Screener tabs, Telegram tips, a spreadsheet, Claude ? Drop it below. I'd love to compare notes on what's working and where things fall through the cracks.
Heym
Thanks for sharing the details here, Shams. The line-level citation approach stands out because most retail investors give up on primary research the moment they hit a 40-page annual report, and having the agent point straight to the filing and line removes the biggest excuse for skipping the homework. A concrete case: someone holding a small-cap position who wants to know the moment a related-party transaction shows up in a new filing, without reading every disclosure themselves.
How does the system handle filings that get restated or amended later, does it flag the earlier version as superseded or just append the update?
I'd also be curious how you're thinking about coverage outside India as you grow, since research quality and filing standards vary a lot by market.
@mbakgun Our engine has a document scoring method which includes the document type and the recency of it. This takes care of any minor discrepancy in reported numbers from the management since we always refer to the latest.
And imo, this is entirely a data indexing game & who can enable context loaded with actionable intelligence at the earliest.
The US already has big players such as Alphasense, Rogo & Bloomberg with deep pockets playing this game. Hence, our focus is primarily on the Indian & larger APAC markets before finding our footing in the US & Europe.
the "every claim linked to the filing" framing is the right instinct for anything touching money. a question on scope though: filings are the paper trail, but a lot of the actual signal in earnings season comes from what management says off-script on the call (tone, evasiveness on a specific analyst question, guidance walked back verbally before it's in writing). do you ingest call transcripts too, or is it filings-only for now? if it's filings-only, that seems like a real gap between "what changed on paper" and "what the company is actually signaling."
@galdayan We go one step better and actually join the call on your behalf and transcribe it entirely and make it available for analysis across Omnisearch, Copilot, Canvas & earnings.
i.e., you can listen to the actual concall split by speaker & sentiment.
@shams_rizvi_kyc that's a much bigger lift than I expected, actually joining the call rather than parsing a transcript after the fact. for the sentiment split, how do you handle management switching between english and hindi mid-answer on a lot of indian concalls, does the sentiment model hold up across that code-switching or is that still rough around the edges
Congratulations on the launch, guys!! The UX is pretty clean :)
The line-level citations are the standout for me. "Never just trust the AI" is exactly the right posture for anything touching money.
Curious about one thing: how often do the agents re-examine an existing thesis? Is it event-driven- a new filing drops and the agent re-checks or on a schedule? Would love to understand how it flags when something you already hold has quietly changed.
@naman_shukla Trust through transparency was one of the key pillars when we embarked on developing this product.
And yes – it's dynamically updated in real time with the latest filing we receive. It checks the recency as well as the document type before updating the thesis. Not all documents are the same.
p.s. You'd be surprised how often a company changes their EPS value in the financial results & then in the subsequent press release. :p
I like that you’re focusing on company research rather than just market trends. Are the insights generated entirely from public information, or can users connect their own research sources as well?
@amjad_shaik At the moment, Only publicly available information.
@shams_rizvi_kyc Thanks for clarifying. That makes sense for an initial version. I’ll be interested to see how the product evolves as teams want to combine public research with their own internal data.
how does the alert system actually decide what counts as material, especially for smaller filings that might still move a position? curious how much i can customize the sensitivity before it just becomes noise
@azad1295492 The idea is to eliminate noise. We do a fairly decent job of not populating your inbox but investing is so subjective and what inbox, be material for one subjective, for another.
We already have all the levers, & the customer will soon have the ability to customise them based on materiality, document type and our scored materiality.
The link-back to filings feature is genuinely useful, made me trust the summaries way more than usual AI research tools. Coverage alerts sound promising too.
@arinburtakpq70 Thanks, Arin! Our aim is to give full visibility to investors as they deploy their precious capital.