Ayan Barua

@cjpedregal do you need a launch partner for your API?

There are some tweets out there saying that Granola is trying to lock down access to your data. Tldr; we are actually trying to become more open, not closed. We’re launching a public API next week to complement our MCP. Read on for context. A couple months ago, we noticed that some folks had reversed engineered our local cache so they could access their meeting data. Our cache was not built for this (it can change at any point), so we launched our MCP to serve this need. The MCP gives full access to your notes and transcripts (all time for paid users, time restricted for free users). MCP usage has exploded since launch, so we felt good about it. A week ago, we updated how we store data in our cache and broke the workarounds. This is on us. Stupidly, we thought we had solved these use cases well enough with our MCP. We’ve now learned that while MCPs are great for connecting to tools like Claude or chatGPT, they don’t meet your needs for agents running locally or for data export / pipeline work. So we’re going to fix this for you ASAP. First, we’ll launch a public API next week to make it easier for you to pull your data. Second, we’ll figure out how to make Granola work better for agents running locally. Whether that’s expanding our MCP, launching a CLI, a local API, etc. The industry is moving quickly here, so we’d appreciate your suggestions. We want Granola data to be accessible and useful wherever you need it. Stay tuned.

@JasonrShuman They led series A in anthropic AND DeepMind 🙇

Which Tier 1 VC led the @AnthropicAI Series A? Bet you couldn’t guess.

Agree — this is important work. There are a bunch of BS companies getting marked up like crazy and they are far more attractive to folks getting into startups for thr first time. This means founders building great companies AND not playing this short term game have to do more work than necessary to convince to go on a mission

@JayaGup10 Glad you’re taking a step at clearing the fud. Valuation isn’t a scoreboard - it’s a coordination tool. If it works for investors but breaks employee upside, you’re borrowing trust from the future. We, founders, own that tradeoff.

Unicorns are everywhere right now. Companies hitting $1B valuations on $2M ARR. Seed-stage startups announcing $500M caps. Every week another AI company crosses the billion-dollar mark. Even 2021 didn't move this fast. If you're evaluating a startup offer right now at one of these companies, there's something you need to understand:

@pk_iv ikea version of love is blind

The meeting room setup at Reinvent is so dystopian. Feels like I’m in the B2B SaaS version of Severance.

Vercel uses Salesforce. Instead of buying an AI SDR, they built a custom one (which is a new category anyways). They use Tray to manage the internal data pipelines. Vercel is not necessarily rebuilding its GTM stack, Vercel is deploying growth engineering to what would traditionally be outsourced to an agency.

Doesn't that presume that your sales process, supported by a CRM, couldn't be a differentiator? Thinking about what vercel is doing - distribution advantages driven by custom software solutions in a space (CRM, RevOps) that could feel like commodities to an outsider. I'd bet there are plenty of startups where "building a custom CRM" isn't the worst idea. x.com/lennysan/statu…

“Every hour spent recreating Salesforce or Workday is an hour not spent building the proprietary capabilities that actually move the business forward.” This is one of the more essential reason why most companies are not going to have AI clone their system of record apps and manage them all themselves. Every project you take on is a tradeoff against some other important project. You “hire” software because someone else has already solved the problem for you. And that problem has a long tail of dependencies, like security, maintenance, system upgrades, and on and on. AI will certainly be used to build competitors in all these categories, in some cases adding pricing pressure, but more likely just expanding the market sizes to reach more businesses via vertical or segment expansion. But what’s most likely is we will use AI to build all the *other* software that doesn’t exist yet. Which will incidentally be about 10X more software than we have already. IT teams will be able to finally go solve the long tail of requests that come in through the business for custom apps, integrations, and automations.

@jatinsandilya_ @lighthousehq_ Congrats Jatin — you were a SF native already in spirit, but this just makes it official

big personal update: as of today I'm officially an extraordinary alien in the US 🇺🇸 huge shoutout to @lighthousehq_ for being an incredible partner throughout the journey, cannot recommend them enough! also big big thanks for everyone who supported along the way @julianweisser @minney_cat @JosephJacks_ @gshewakr @beondeck @MatthiasWagner @OSSCapital @Loganullyott @ayanb @laurenzlong so excited to be building the next phase of my life in the land of opportunity. more soon & back to work!

@tod @pipedream @Workday congrats @tod and team! Huge news

It’s a big day here at @Pipedream – we’ve entered into a definitive agreement to be acquired by @Workday. We are combining our AI products and connectors with Workday's enterprise customer base and 75M+ users under contract, and will integrate our technology across Workday's agent development stack.

@Hadley it is boiling down to the team so series A is more similar to seed now

Series A and B markets are strange right now. It is harder than ever to raise, but the few companies that do are raising very large rounds at higher valuations. A and B volume is way down while median valuations are way up. This would seem to indicate the market is getting better at picking winners. I do not think that is true. Picking winners in AI is harder today. Moats are weaker, switching costs are lower, and most teams build on the same base models. What is really happening is capital concentration. Funds have a lot of dry powder and very few companies they feel confident backing, so they pile big dollars into the small group that builds early consensus. It looks like conviction, but it is really scarcity and FOMO. Capital can help, but it is not a moat. The companies breaking through the filter have more leverage than ever. The ones that do not may simply be on the wrong side of a distorted market, not lacking long term potential.

@austinh___ a lot of different ICPs in there, hence harder to nail down the product form factor

Most teams juggle a data warehouse, a transformation tool, and a CRM. Imagine if those boundaries blurred-until the CRM itself became obsolete. That’s the direction we’re heading, and it’s way overdue.

@hassaanrza Epic timing on the “we have .. Salesforce”

The interface of the future is human. We’ve raised a $40M Series B from CRV, Scale, Sequoia, and YC to teach machines the art of being human, so that using a computer feels like talking to a friend or a coworker. And today, I’m excited for y’all to meet the PALs: a new human-computing interface. PALs are emotionally intelligent, multimodal, and capable of understanding and perceiving. They can see, hear, reason, and even look like us. We’re releasing our 5 favorite PALs to start. Each PAL has its own distinct personality- from AI assistants to best friends. PALs: - Meet us where we are. Face-to-face over video call, on the phone, or even by text. - Are always thinking. They’re proactive, reach out first, remind you about what you forgot, or might just check in on you. - Understand us, finally. PALs can see us, understand our tone, emotion, and intent, and communicate in ways that feel more human. - Evolve with you. PALs have advanced memory, remember your preferences and needs, and adapt themselves over time. - Are capable. PALs can handle complex tasks — from responding to your emails to moving your schedule around to creating docs and doing research for you. Science fiction promised us a new human-computer interface, beyond the GUIs of yesterday, a human-like interface that would feel second-nature to use. That future never came, until now. Charlie’s story brings this idea to life. We’re excited for you to meet Charlie and his PALs for free at tavus.io Enjoy the film 👇

@jaltma the best ones are, but they may not be able to signal that publicly 100% of the time?

Most VCs should be much less sure of most of their opinions most of the time.

@calcsam everything is a prologue until it isn't

gatsby was just the prologue

@kwharrison13 This is a good chatGPT Atlas use case

We have a company doing 8-figures of revenue, growing 10x, signing 6-figure deals with massive enterprises. That company’s traction struggles to catch VC attention cause the logos are all multi-billion companies but in a less well-known category. None are VC backed.

@auren 0 to 100M arr on day 1

New billion dollar business idea: SecondChew

@Hadley it all evens out in a sufficiently long time horizon, because customers only care about their own problems

We’re in a weird era for startups where perception has become the moat. Raise early, get labeled a “category leader,” and that perception attracts capital, customers, and talent until it’s no longer perception. The irony is many kings were crowned long before they’d earned it.

@startupandrew @taskletai Congrats Andrew!!

Today we're launching Tasklet — an AI agent for automating your business. Unlike ChatGPT, @TaskletAI actually does the work for you: connecting to your tools, triggering automatically, and handling tasks while you sleep.

solid advice

@ankrgyl eventually all these internal departments run into hairy implementation, integration, support, customization problems which is great - paves the way for SaaS 3.0 (or whatever version we are on for agentic B2B)

There appears to be a gold rush for convincing internal departments at enterprise companies to build agents on your platform. Similar to the early days of no code / low code builders. Interesting tradeoffs between lock in, UX, cost, etc

A true CRM sync for a SaaS company is usually a lot of API-based code and config, and that is really hard to vibe code because it needs three types of context: - SaaS companies' codebase (often large) and schema (drifting) - Customer of SaaS companies configuration (every Salesforce org is different) - Salesforce APIs themselves are changing often, and the approach around error handling, rate limiting, pagination, etc needs a ton of work What can be vibe code is a thin AppExchange app built up for reporting, so there is _some_ merit to this. They cannot make it too good though, it will kill a bunch of consulting companies that charge by the hour for creaeting janky SFDC apps

absolutely incredible that Salesforce is really calling this Agentforce Vibes and the agent within it Vibe Codey

@samlambert ragebait marketing works