Kevin Lynch

building

[Kevin Lynch @klynch]

Venture capital is all about figuring out which questions are the right questions to ask, and since we don’t have a clue what the right answer is, we’re very interested in the process by which the entrepreneur get to the conclusion.

- Don Valentine

[Kevin Lynch @klynch]

AI customer support is a tricky technical problem. You need to make agents open / agile to do useful things, but control them so they don’t go off the rails at scale (Meta’s bot causing a massive security incident, Amazon + many others’ serving content unrelated to a company’s customer support, etc)

[Addy Osmani @addyosmani]

https://t.co/0DyFIXJueI

[Kevin Lynch @klynch]

@nikitabier @IterIntellectus I like the new design

[Kevin Lynch @klynch]

@jordihays @samsheffer This is waymo exciting

[Eoghan McCabe @eoghan]

We’re launching our next major new product live from SF tonight. I think this one will be interesting even for those with no need for Fin. Will be a great demo of what the future of work software will look like.

Stream starts at 6.15pm PT.

fin DOT ai SLASH hello

[Aaron Levie @levie]

Whether it’s existing consulting firms, new ones that emerge, FDEs from agent vendors, or new internal agent engineering roles, the amount of work that is going to be created to implement agents in enterprises will exceed anything we imagine today.

The complexity of implementing agents in any existing organizations is very real. When I talk to large enterprises, as you move from a chat paradigm to agents that participate in meaningful workflows, there are a number of things they need to do.

First, you have to get agents to be able to talk to your data securely across your systems. In many cases, enterprises have decades of legacy infrastructure that contain the valuable context for AI agents. That’s going to take a ton of work to go modernize and move to systems that work well with agents.

Then, you need to ensure that you’ve implemented agents with the right access controls and entitlements, the right scopes to be safely used, and have ways of monitoring, logging, and securing the work that they do.

Next, you need to actually document the processes in the organization in a way that agents can utilize for doing the work. You also need to figure out what the new workflow looks like when agents and people are working together on a process, and who steps in where. Just replicating the old workflow will mute the gains. Oh and you likely need to create evals for your top new end-state processes.

Finally, you have to keep up with a rapidly changing set of best practices and architectural shifts happening in the agent space. While it’s fun for people to change their personal productivity tools on a dime, it’s 100X harder to do this in a business process. The speed of change is a blessing and a curse right now for anyone trying to keep a stable system design.

All of this means that individuals and companies that develop expertise on the above set of components (and more) are going to be needed to help organizations actually implement agents at scale. This is also the rationale for vertical AI agents right now that can go in deep on a business domain and help bring automation to it.

This is a huge opportunity right now whether you’re doing this internally or as an external business provider.

[Maddie Hall @maddiehalla]

1/ @OctopusEnergy is backing Living Carbon with $500M to reforest degraded land and remove CO₂ across North America. This major project financing comes with an additional ~$13M investment in our carbon business.

Living Carbon is putting low quality land back to work. Our innovative reforestation turns low quality land into thriving forests faster to remove carbon or produce sustainable forest products.

Read more about it in today’s @WSJ (link in the comments) →

[Darragh Curran @darraghcurran]

9 months ago we publicly committed to 2x the productivity of our R&D org at @intercom.

It was scary. It wasn't always clear we'd pull it off.

We hit it with 3 months to spare. In fact, looking back 16 months - we've 3x'd.

Here's what actually happened (with receipts): 🧵

[Brian Scanlan @brian_scanlan]

This is no longer an experiment.

Announcing the Fin CLI - the agent-first way of signing up to, configuring and using Fin and the Intercom helpdesk. Ask your agent to use fin dot ai slash cli to set Fin up from scratch.

[Brian Scanlan @brian_scanlan]

Just like everybody else this week, we've also been experimenting with cli interfaces.

You can now sign-up for Intercom, configure and install Fin and the messenger from inside your agent.

npm install -g @intercom/cli

It's wild how fast things are changing.

[Tomasz Tunguz @ttunguz]

Would you choose one software over another because it has a proprietary model with better performance?

Two companies shipped custom AI models today (three in a week counting Cursor!), raising that question. Intercom launched Apex 1.0, a model for answering customer support tickets. Chroma released Context-1, a model for multi-hop agent search.

Apex 1.0 beats GPT-5.4 & Claude Opus 4.5 on customer service tasks. Context-1 scores 97% on agent search benchmarks. One Intercom gaming customer saw resolution rates jump from 68% to 75%.

History suggests these gains may be temporary. As general-purpose models improve, today's specialized advantage erodes. But with GPU shortages, inference costs will spike, perhaps this will be the moment for built-for-purpose more efficient models.

Intercom built Apex to differentiate in a competitive market. Chroma's bet is different. Context-1 is open-source under Apache 2.0. Anyone can use it. The model isn't the product. It's marketing rather than sales. Distribution & brand building for their vector database infrastructure.

Two philosophies. Proprietary model as differentiation versus open-source model as adoption mechanism.

"As features become ~free to build, the technology factors that will differentiate the players will be the AI under the hood. If you're using the same general-purpose off-the-shelf model as everyone else, you have no durable differentiation." - Eoghan McCabe

AI models offered by software vendors have become a new axis upon which to compete. In the marketing arena, models drive attention & distribution. At the bottom of the sales funnel, they serve as competitive differentiators in performance.

https://t.co/dB8l0gocFz

[Fergal Reid @fergal_reid]

https://t.co/niV386N65m

[Eoghan McCabe @eoghan]

https://t.co/I2Hf40EpIt

[Brian Scanlan @brian_scanlan]

We've been building an internal Claude Code plugin system at Intercom with 13 plugins, 100+ skills, and hooks that turn Claude into a full-stack engineering platform. Lots done, more to do. Here's a thread of some highlights.

[Eoghan McCabe @eoghan]

I wrote a short post to explain why exactly we raised new capital and what's about to come next. And as a bonus, talk about the unconventional way we raised it.

[Eoghan McCabe @eoghan]

AI is coming for all of our jobs. But only the most bloated of organizations can be cut by 40% with it today.

[Kevin Lynch @klynch]

@jaminball @altcap @TrumpAccounts @InvestAmerica24 Well deserved!

[Matthieu Hafemeister @matt_haf]

Excited to share that @Concourse_ai has raised a $12M Series A led by @Standard_Cap with participation from @a16z, @CRV, and @ycombinator to fundamentally change how finance teams work with AI agents.

Concourse builds AI agents specifically for finance teams. Instead of spending 75% of their time on manual work, teams who use Concourse now produce 6x more analysis and create presentations in minutes.

Over the last 12 months we’ve grown revenue by 19x, customer logs by 13x, and are now deployed inside large enterprises like Front, Tecovas, and Palo Alto Networks.

Alongside our fundraise, today we’re making Concourse available generally so teams of all sizes can connect their data and start using Concourse AI agents in minutes.

[TBPN @tbpn]

Sequoia's @GuptaRK22 and @gradypb on potential layoffs and why the AI doomers are wrong:

"Technology is neither good nor bad, it's what we do with it. The same technology that powers a nuclear bomb can power a nuclear power plant."

"A lot of what people are doing when they go and talk about the technology is they talk about all the things that might happen. I would rather us talk about the fact that it's going to get more and more powerful, and it's more and more incumbent on all of us to make sure we use it for something great."

"If I could give my kids one thing, it wouldn't be kindness. It wouldn't be curiosity. It would be agency. The technology is going to be what all of us make of it."

[Kevin Lynch @klynch]

@kane @gregmcbeth @adamhousley @GavinNewsom Time to add a new chapter: https://t.co/AlKCOBd5qG