Realtime AI meeting assistant starting with sales
TL;DR
Fern is a realtime AI that joins your meetings as either an invisible copilot or an active participant– your choice! It realtime proactively coaches you through every conversation so you always show up like a top 1% professional.
Hey everyone, we're Amit and Robert, cofounders of Ishiki Labs.
The Problem
“AI meeting assistant” has become synonymous with a note-taker waiting to summarize at the end. That's not a meeting assistant. That's a fancier transcript.
Building a realtime AI meeting copilot is hard in two distinct ways.
The invisible assistant must understand the live meeting context at blazing speed and surface the right thing at the right moment – questions to ask, things you're missing, who you're talking to. Timing is everything. Current AI is purely reactive. Making AI an active participant in discussion is even harder. AI has no social awareness – every pause becomes a prompt, every gap an invitation to speak.
AI that is socially aware and proactive is an unsolved problem. That's what we're building.
The cost of this gap is real. Cofounders, AEs, recruiters, investors – in discovery calls, sales meetings, investor conversations. The moments that matter most, and AI is absent from all of them.
How Fern Solves it
Fern joins your meetings as a true participant or an invisible copilot – your choice. It has two modes:
Full Presence: Fern joins visibly, listens, and speaks up when invited. It offers realtime opinions, deep research, and can complete tasks like filing notes or sending emails on the spot. Built for internal product, engineering, and strategy meetings. Watch the demo- https://www.youtube.com/watch?v=vRJfZDUHvSY
Shadow Mode: Silent and invisible to everyone else. You get a private realtime feed: live summaries, questions to ask, things you're missing, context on who you're talking to. A sharp colleague whispering in your ear – with fully customizable insights. Built for client calls, investor meetings, and any conversation where you want invisible support.
Both modes include a post-meeting summary, transcript, and full insights.
Team
We're both ex-Meta. Amit is an AI Scientist from LLaMA (now Meta Super Intelligence Lab) and Reality Labs – PhD in AI, 20+ publications at NeurIPS/CVPR, trained multimodal assistants for smart glasses. Robert was systems and infra lead in Reality Labs, with four years building low-latency multimodal systems at Meta and Citadel Securities.
Our mission: make AI socially timed and contextually aware – so every conversation you have runs like a top 1% professional.
Our Ask
Download Fern and if you're a founder or AE who lives in high-stakes conversations and don’t want to settle for AI note-takers, reach out: team@ishikilabs.ai
How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?)
Making multimodal AI truly human-like is hard. We are a highly technical team with deep research and system experience on building multimodal assistants. Amit has a PhD in AI, worked as an AI research scientist at Meta, and trained multimodal assistant for smart glasses. He also researched with Meta SuperIntelligence Lab training multimodal LLMs and has 20+ publications at top conferences in AI including NeurIPS, CVPR, and ICASSP. For the last 4 years, Robert has been building advanced orchestration systems for running multimodal assistants on smart glasses, optimizing for latency and compute. Before Meta, Robert worked on developing research infrastructure at Citadel Securities.
We originally built socially conscious AI - an AI that knows when to talk and when to stay silent. But as we talked to more customers, it became clear that sales is the industry that requires social intelligence the most - as how well you talk directly correlates to how many deals you close. And so we’re applying our research and work on social consciousness into building Fern, a realtime sales copilot and intelligence platform, to help your team sell better.
What is your long-term vision? If you truly succeed, what will be different about the world?
World will have a human like multimodal AI assistant. Simplest way to say that is: if an expert human watching a video stream can help, we're building the AI that can too.
