dharmesh

23.6K posts

dharmesh banner
dharmesh

dharmesh

@dharmesh

Co-founder/CTO, HubSpot ($HUBS). Mission: Help millions grow better. Publish https://t.co/sgeuiJi7wU newsletter (2M+ subscribers). Builder: https://t.co/xirHeCKZPl

Boston, MA Katılım Mart 2008
772 Takip Edilen354.2K Takipçiler
Sabitlenmiş Tweet
dharmesh
dharmesh@dharmesh·
2023 was the year of chat UX powered by generative AI. That led to me hacking on ChatSpot and launching it last year. 2024 is the year of Agent AI, so that's what I've been hacking away late nights on. I'm obsessed with #AgentAI. Was up past 3am last night. I have a handful of simple (but useful) agents that I built for myself and one I built for my wife. Will start releasing them one at a time. I'm using a bunch of LLMs (GPT-4, Claude, Gemini) and a bunch of proprietary/paid data sources. Will be fun to see the bills rack up. :) Releasing it all for free. You can join the waitlist at agent.ai Thanks for your support.
dharmesh tweet media
English
170
204
2.2K
644.3K
dharmesh retweetledi
Aaron Levie
Aaron Levie@levie·
One of the biggest lessons thus far in building AI agents is you have to be brutally unsentimental in your architecture. The models get better and better at handling things you previously built scaffolding for, you need to ruthlessly jettison your prior tech to get those new performance gains. The rough loop of building AI agents looks something like: 1. Build a bunch of systems around the LLM to ensure that the agent can solve specific tasks very well 2. The model capabilities dramatically improve, rendering many of those systems redundant or even harmful 3. Remove prior scaffolding to get the new performance gains from the agent 4. New capabilities emerge in the models that let you solve a new set of much harder problems 5. Go back to step 1 For instance, in our new Box Agent, from the moment we designed the original architecture to the ultimate release, we had to evolve multiple components of agent harness simply because some parts were creating unnecessary constraints for the agents as models improved. The models continued to get insanely good at more complex reasoning, improvements in using search and other tools, writing code on the fly for new capabilities, improving context window performance for accuracy, and more. Many of the mitigations we put in place for the Box Agent (like to appropriately find data that users were looking for, or ways of chunking text to deal with context window limitations), eventually meat we got lower quality results or meant we were overfitting for specific use-cases, as soon as the models got better. The main lesson is always make sure you’re taking advantage of the frontier capabilities and don’t become nostalgic around the tech you’ve already built.
English
63
40
394
55.2K
dharmesh retweetledi
Brian Halligan
Brian Halligan@bhalligan·
If Steve Jobs were still alive, I bet Apple would be at a $10trillion market.
English
52
28
481
86.4K
dharmesh retweetledi
Michael Cyger
Michael Cyger@MichaelCyger·
True words from @Dharmesh. I failed over and over again last week getting OpenClaw up and running, and then when I did I failed getting Mission Control up and running. I will most definitely try again in a few months.
Michael Cyger tweet media
English
1
1
10
2.2K
dharmesh retweetledi
Dan Wolchonok
Dan Wolchonok@danwolch·
Today's meeting tools give you a transcript, but nobody tells you how you did. I'm so excited to announce Work Coach (a new mac meetings app). Work Coach helps you by: - Observing you in meetings (interviews, 1:1s, sales calls, team meetings) - Analyzing how you show up to those meetings - Identifying opportunities to improve - Replaying the exact moments from your meetings where it happened - Role play conversations (intros in an interview, asking for a raise, etc) - Interviewing your coworkers to get feedback that you're not getting I'll put the download link below.
English
15
12
47
13K
dharmesh retweetledi
Brian Halligan
Brian Halligan@bhalligan·
What Jack Dorsey learned from Bill Walsh (49er’s legend) about going from startup founder to scaleup ceo.
Startup Archive@StartupArchive_

Jack Dorsey on what he learned about culture from 49ers coach Bill Walsh When Bill Walsh joined the 49ers, they were the worst team in the NFL. Within three years, they were Super Bowl Champions. In his book, The Score Takes Care of Itself, he writes: “Winners act like winners before they’re winners…The culture precedes positive results. It doesn’t get tacked on as an afterthought on your way to the victory stand. Champions behave like champions before they’re champions; they have a winning standard of performance before their winners.” And he provides six guidelines for establishing a standard of performance: 1. Start with a comprehensive recognition of reverence for and identification of the specific actions and attitudes relevant to your team’s performance and production. 2. Be clarion clear in communicating your expectation of high effort and execution of your Standard of Performance. Like water, many decent individuals will seek lower ground if left to their own inclinations. In most cases you are the one who inspires and demands they go upward rather than settle for the comfort of doing what comes easily. Push them beyond their comfort zone; expect them to give extra effort. 3. Let all know that you expect them to possess the highest level of expertise in their area of responsibility. 4. Beyond standards and methodology, teach your beliefs, values, and philosophy. An organism is not an inanimate object. It is a living organism that you must nurture, guide, and strengthen. 5. Teach “connection and extension.” An organization filled with individuals who are “independent contractors” unattached to one another is a team with little interior cohesion and strength. 6. Make the expectations and metrics of competence that you demand in action and attitudes from personnel the new reality of your organization. You must provide the model for that new standard in your own actions and attitude. Twitter and Square founder Jack Dorsey encourages anyone thinking about leading teams or building a company to read this book: “What’s important about this is that as you start building a team, you need to set expectations around how people need to perform in the company—how people need to act in the company. And these can be very simple things, but without that, you are rutterless—you will react to the outside. And if you react to the outside, you are building someone else’s roadmap and you’re building someone else’s dream instead of your own.” Video source: @ycombinator (2013)

English
7
2
42
20K
dharmesh
dharmesh@dharmesh·
Love this framing -- and could not agree more. (Disclosure: I'm an investor in some of these startups). Agents are going to need some core tools and capabilities to do work. It's why HubSpot is building an Agentic Customer Platform to provide the CRM and GTM context agents will need (both ours and others).
Shiv@shivsakhuja

Lots of companies are now building primitives for an economy where AI agents are the primary users instead of humans. They're betting on an economy of AI coworkers. 1. AgentMail (@agentmail): so agents can have email accounts 2. AgentPhone (@tryagentphone): so agents can have phone numbers 3. Kapso (@andresmatte): so agents can have WhatsApp phone numbers 4. Daytona (@daytonaio) / E2B (@e2b): so agents can have their own computers 5. Browserbase (@browserbase) / Browser Use (@browser_use) / Hyperbrowser (@hyperbrowser): so agents can use web browsers 6. Firecrawl (@firecrawl): so agents can crawl the web without a browser 7. Mem0 (@mem0ai): so agents can remember things 8. Kite (@GoKiteAI) / Sponge (@PayspongeLabs) : so agents can pay for things. 9. Composio (@composio): so agents can use your SaaS tools 10. Orthogonal (@orthogonal_sh) so agents can access APIs easily 11. ElevenLabs (@ElevenLabs) / Vapi (@Vapi_AI) so agents can have a voice 12. Sixtyfour (@sixtyfourai) so agents can search for people and companies. 13. Exa (@ExaAILabs): so agents can search the web (Google doesn’t work for agents) If you stitch all of these together, you get a digital coworker that looks more human than AI.

English
26
12
143
48.3K
Rohan
Rohan@rohanagrwl·
We have a @HubSpot sub for months but none of the dummies in sales used it as its 'too complicated' Played around with its AI Agent & it actually gets stuff done instead of having to learn how any module/functions work @dharmesh your team cooked, this saves a ton of time 🙌🏻
Rohan tweet media
English
1
0
1
426
dharmesh
dharmesh@dharmesh·
@btaylor Nice! Loved the subtle "haunted" easter egg in the messages shown while it's working (keeping with the "ghost" theme).
English
0
0
12
3.6K
Bret Taylor
Bret Taylor@btaylor·
Today, Sierra is releasing Ghostwriter, our agent for building agents. With Ghostwriter, you can create an AI agent for your customer experience — one that can chat, pick up the phone, speak dozens of languages, take action on your systems of record, and be protected with industry-leading guardrails — simply by having a conversation. No clicking, no forms, no menus. Codex and Claude Code have transformed how we build software, making it possible for software engineers to orchestrate and review the work rather than doing all the work themselves. We think the same transformation will happen for all software. Rather than every enterprise app having a web app for humans and an API for automation, every software platform’s UI will be an agent that can do the work on your behalf. I recorded a demo of my building and optimizing an agent with Ghostwriter so you can see how powerful and easy it is to use. It’s completely changed the way our early adopters build agents, and it’s changed the way I think about the software industry. Let me know what you think, and, if you’re interested in trying it out at your business, please reach out directly.
English
165
302
3.2K
1M
dharmesh retweetledi
Jason Cohen
Jason Cohen@asmartbear·
ICYMI: Remastered classic article: Customer love doesn't come from feature bullet points. That’s utility. Utility is very very good. It’s just not Love. Love is something deeper, less incremental, less data-driven, more gut feel, more emotional. longform.asmartbear.com/creativity-ove…
English
4
3
8
2.5K
dharmesh retweetledi
adarsh
adarsh@adarsh_exe·
Traditional coding benchmarks do not reflect how software is actually built and maintained. That's why we built a new benchmark, APEX-SWE, in partnership with @cognition. It measures whether AI models can perform complex, real-world software engineering work to ship systems that work and debug them when they don't. @OpenAI GPT 5.3 Codex (High) tops the leaderboard at 41.5% on Pass@1.
English
124
155
849
191.5K
dharmesh retweetledi
Bret Taylor
Bret Taylor@btaylor·
I grew up hacking on and building with early open source software - Linux, Apache, GCC. For me, a big part of the culture of open source was people making their software their own. I have been thinking about open source in the age of coding agents. There is one world where many open source projects become less important because we all independently recreate the functionality with coding agents. Is there another where open source platforms create an ecosystem with their harness and agentic hackability, like OpenClaw has done? If an open source database or content management system shipped with an exceptional agent harness that made easy to extend and improve with agents, would that make individual developers more inclined to start there rather than start from scratch? Likewise, in the spirit of the hacker culture I grew up with, will power users like me start to seek out systems that are designed to be hackable with agents? I’d love a desktop operating system that was more hackable by casual users like me - not just the gratuitous window dressing of the early Linux desktop movement, but set of composable graphics and OS abstractions designed to be composed, mutated, and rearranged by an agent. After 25+ years, maybe this really is the year of Linux on the desktop 😏
English
39
13
178
23.7K
dharmesh
dharmesh@dharmesh·
Following up on this. Evidently, we announced this service was being discontinued last year and it was decommissioned in December, 2025. Details here: developers.hubspot.com/changelog/upco… We have been recommending to anyone still using the old Clearbit logo service to switch to logo.dev.
English
2
0
7
2.3K
@levelsio
@levelsio@levelsio·
🎲 Clearbit just nuked their free logo service You could call it by for example: > logo.clearbit .com/microsoft .com To get Microsoft's logo What's sad is they didn't just 301 redirect it to another service, like Google, which means lots of sites that rely on it to show logos of companies (like many of my sites) now break They got acquired by Hubspot, so a cool fix for @dharmesh that I DM'd already would be to 301 it to Google's logo API: > https://s2.googleusercontent .com/s2/favicons?domain=microsoft .com&sz=128 Or DuckDuckGo's logo API: > https://icons.duckduckgo .com/ip3/microsoft.com.ico I love websites that offer a service and then when they abandon it, respect the web and somehow keep it going or 301 redirect it to keep the web functioning, because when you browse older sites on the web, everything is just broken because nobody ever maintains shit after they sell the company!
@levelsio tweet media
English
62
16
721
109.6K
dharmesh
dharmesh@dharmesh·
@levelsio Thanks for the heads-up on this. Sorry for the lag in response time. Was on the road this evening. Will have the team look into this in the morning.
English
3
0
33
3.4K
dharmesh retweetledi
Brian Halligan
Brian Halligan@bhalligan·
"Your job is process design, my job is outcome delivery" This might be my favorite single clip of the podcast to date:
English
13
15
169
25.8K
dharmesh
dharmesh@dharmesh·
My utmost respect to those amongst you that are strong enough to click the "Tidy Up" button when Slack tells you that you have channels you haven't visited in a while. It is a state of being I aspire to some day.
dharmesh tweet media
English
7
0
14
2.2K
dharmesh retweetledi
Paul Graham
Paul Graham@paulg·
Someone asked what advice founders ignore. That they: 1. Should change their name. 2. Should launch fast. 3. Shouldn't treat fundraising as success. 4. Shouldn't assume they can raise because it's time to. 5. Should fire bad people quickly. 6. Shouldn't talk to acquirers.
English
313
246
4.3K
912.7K
dharmesh retweetledi
Brian Halligan
Brian Halligan@bhalligan·
Old Marketing: $ ➡️ ads ➡️ Prospects New Marketing: $ ➡️ tokens ➡️ Prospects
English
17
2
64
11K
dharmesh retweetledi
Garuba Richard
Garuba Richard@anodeen·
HubCode bringing true no-terminal vibe coding straight into the HubSpot UI, with full agent.ai reuse (2k+ agents), screenshot prompts, and real data access. This feels like the natural next evolution. Cloud-only with private beta perfect for rapid iteration. Already heading to hubcode.com to apply. Builders are eating good.
English
1
2
2
2K