Teahose.
SIGN IN
NEW HERE — WHAT TEAHOSE DOES
We read the entire AI & tech firehose — so you don't have to.
PODPodcastsAll-In, No Priors, Acquired…
NEWNewslettersStratechery, Newcomer…
PAPPapersPhysical AI research
PHProduct Huntdaily launches
VCInvestor ScoutSequoia, a16z, Benchmark…
CLAUDE DISTILLS →
7 reads, 30 sec each — free, 6 AM ET.
+ a live graph of the companies, people & themes underneath.
HOME/THE A16Z SHOW/The Professional Network Built f…
POD
// EPISODE
THE A16Z SHOW

The Professional Network Built for You and Your AI Agent

DATE November 21, 2026SOURCE THE A16Z SHOWPARTICIPANTS BELLA NAZARI, DAVID BOOTH, SOPHIA DEW
// KEY TAKEAWAYS6 ITEMS
  1. 01Talent Isn't Scarce, It's Illegible
  2. 02Reputation Over Resume: "The Work They've Done and Who Can Vouch for It"
  3. 03The Most Impactful People Are Often the Least Visible
  4. 04AI Agents Are Becoming First-Class Network Participants
  5. 05Curation (Lists) as the Layer on Top of the Graph
  6. 06Awards as a Network-Seeding Mechanism, Not a Prize
In this episode

1. Key Themes

Talent Isn't Scarce, It's Illegible

The core problem Cosign targets is a legibility problem rather than a supply problem. Static profiles and resumes can't show who actually did the work or who is available at the moment a founder is looking. David Booth (leads Ecosystem at a16z) argues that "there is no shortage of talent. The way I look at it, it's actually very illegible to see who is the right person for the right role. The coincidence of wants of the right person being available right now at the same time as the potential co-founder or the company that's trying to hire them." [00:04:47] He adds that "there is only so much that you can tell from what people post online on LinkedIn. Only so much that you can read in profile and it's sort of a more static profile." [00:04:47] Sophia Dew framed the other side of the market: founders struggle to find talent while talented people "struggle to find job opportunities or I'm unhappy with my current job." [00:04:26]

Reputation Over Resume: "The Work They've Done and Who Can Vouch for It"

Cosign is built on vouching as the primary signal. Booth: "the thing that we've always cared about is like what's the work they've done and who can vouch for that work? And that's the root of Cosign for us." [00:05:15] Bella Nazari (growth lead for Cosign, formerly on the Speedrun team) emphasizes that the "cosign" is a new primitive: "recommendations have existed, but like not in the form of like a cosign." [00:07:05] The action is explicit and public, so the endorsement carries information a resume can't.

The Most Impactful People Are Often the Least Visible

Booth argues that real impact inside organizations is poorly correlated with org-chart position or public visibility: "so often the people who are the most impactful in the organizations today, all around us, aren't the ones who are like sitting on top of the org chart." [00:05:15] The unlock is senior validation of rising talent: "when a VP of engineering co-signs the young up-and-coming IC engineer or it's the people who have the product instinct and the ability to ship the product that they've been able to envisage. And that's some of the people that we're now seeing rise up through this." [00:05:44] He also notes that "some of the most visible, most impactful people of the past five years are no longer the ones that are really driving things today." [00:05:44]

AI Agents Are Becoming First-Class Network Participants

The most forward-looking thread: Cosign exposes an agent-accessible interface (an MCP server, per the speakers). Booth: "Some of the most interesting emergent behavior on Cosign we're seeing today is people have connected their agents to it through the MTS, cosign.co forward slash agents. And you can give it to a GrokBot, you can tell it to go out there and find people that you should know. You can tell it, check every day, you know, check the feed every day for things or for people or for anytime somebody that I'm connected to creates a new list, see who's on the list." [00:00:00] This turns the network from a place humans browse into a continuously running discovery layer.

Curation (Lists) as the Layer on Top of the Graph

Booth describes lists as the curation primitive that makes the graph searchable: "if Cosign is this, your profile is across of a very broad subset of people building the most impactful companies at the frontier of technology today, lists and the Cosign Awards is ultimately, it's lists, it's like curation that sits on top of that. And that's what helps, you know, us find an interesting person for you to connect with or for your agent to find and connect with you." [00:00:21] The awards are therefore best seen as a data-generation exercise for a curated, vouched-for talent map.

Awards as a Network-Seeding Mechanism, Not a Prize

The awards (49 awards across 18 people categories, 13 company categories, and 18 product categories, with 21 judges) are explicitly not about winning. Booth: "this isn't a lead award. There's not going to be a gold medal handed out to the person with most votes. It's not actually about votes. It's about identifying a pool of people." [00:10:55] The judges then select "five to ten people across each of the categories." [00:10:55] Booth also notes that for many nominees "they don't need another award or a badge or a shiny medal"; the point is gathering them. [00:05:44]

Communities Around Emerging Roles

Nazari describes a repeatable playbook for the team: "creating these communities around certain pieces of leadership. So it can be like leadership in the talent sector, in the growth sector. It can be engineering. We talk a lot about with us like the rise of the FDE [forward deployed engineer]... what are the broad-based themes of where we're seeing like roles go, work go, you know, companies go. And then putting those people in a room, having those discussions, making friends, making community." [00:12:30] She calls Cosign "a community product, I think, at its root." [00:12:58]

Patience: Network Effects Compound Over Years

Booth frames the value of gatherings as long-horizon: "maybe five years from now, there's a company built by people that met at this thing... you have to be very patient. It takes a long time to play out." [00:13:23] He anchors it in precedent: "the people who met each other while working at PayPal and went on to start, or the people who met each other while working at Palantir and went on to start the next great..." [00:13:44]

Distribution Through the a16z Ecosystem, Then Gifted Outward

Nazari describes the go-to-market logic: "I think a16z is uniquely positioned to start something like this, but then gift it to the rest of the ecosystem. And have all of San Francisco and then maybe more broadly, the world... take on this kind of network and adopt it as their own." [00:03:23]

2. Contrarian Perspectives

The Talent Shortage Is a Myth; the Problem Is Legibility

Most founders and recruiters say they can't find good people. Booth rejects the premise: "everybody is always constantly seeking great talent and there is no shortage of talent... it's actually very illegible to see who is the right person for the right role." [00:04:47] If true, the investable opportunity is in signal infrastructure and trust graphs rather than sourcing volume.

Org-Chart Leaders Aren't Who's Driving Outcomes

Conventional networking and recruiting orient around titles and well-known names. Booth argues the opposite: "aren't the ones who are like sitting on top of the org chart," and "some of the most visible, most impactful people of the past five years are no longer the ones that are really driving things today." [00:05:44] This implies that title-based search systematically misses the people building products.

An Awards Program That Explicitly Disclaims Winning

Awards usually exist to confer status. Here, Booth says the votes are a mechanism, not the point: "It's not actually about votes. It's about identifying a pool of people." [00:10:55] And "it's not designed to be... like a prize giving. It's more like, let's just take the opportunity to get a group of incredible people together, get on the ground floor of the next great talent mafia." [00:11:52] The contrarian bet is that the curated pool, not the trophy, is the asset.

The Network of the Future May Be Primarily Consumed by Agents

Instead of designing for humans scrolling a feed, Cosign is already seeing users delegate network monitoring to bots that check the feed daily. Booth: "you can tell it to go out there and find people that you should know." [00:00:00] That suggests professional networks may compete on machine-readable reputation quality rather than UI engagement.

3. Companies Identified

Cosign

A new professional network built around vouching ("cosigning") people and curated lists, incubated within the a16z ecosystem. Mentioned as the central subject. Booth: "the thing that we've always cared about is like what's the work they've done and who can vouch for that work? And that's the root of Cosign for us." [00:05:15] It exposes an agent endpoint: "people have connected their agents to it through the MTS, cosign.co forward slash agents." [00:14:59]

a16z (Andreessen Horowitz)

Venture firm that runs the ecosystem, Speedrun, and new media teams behind Cosign. Nazari: "I think like A16Z is uniquely positioned to start something like this, but then gift it to the rest of the ecosystem." [00:03:23]

Speedrun (a16z)

a16z's "inception fund" for early-stage founders, where Nazari worked on talent. Nazari: "Speedrun is the inception fund for A16Z. So, you know, if you're looking to get started with your founding journey, it's a good place to start." [00:02:27] She notes the team "think[s] a lot about networks" and grows talent networks. [00:02:58]

MTS (the show/stream host)

The live show hosting the conversation and co-hosting the Cosign Awards party. Sophia Dew: "there's a big party that we're going to help host with you guys in a few weeks where we can finally share all of the results of these awards." [00:01:34]

PayPal

Cited as a historical example of a company whose alumni network seeded later companies. Booth: "the people who met each other while working at PayPal and went on to start..." [00:13:44]

Palantir

Cited alongside PayPal as an origin point for founder networks. Booth: "the people who met each other while working at Palantir and went on to start the next great..." [00:13:44]

LinkedIn

Implicit incumbent and contrast case. Booth: "There is only so much that you can tell from what people post online on LinkedIn." [00:04:47] Dew also joked that the platform surfaced "that old photo of me from LinkedIn." [00:08:41]

Grok (xAI)

Mentioned as an example of an agent a user can connect to Cosign. Booth: "you can give it to a GrokBot, you can tell it to go out there and find people that you should know." [00:00:00]

Evondeck / a16z Growth Engineer Fellows (programs)

Earlier ecosystem efforts Booth has been involved in. Booth on his background: "having been doing this kind of thing for now for five, 10 years across Evondeck and now A6NZ." [00:14:13] On the growth category: "There's a lot of the A16D growth engineer fellows that we've had in the past." [00:10:07]

4. People Identified

David Booth

Leads Ecosystem at a16z; guest and architect of the Cosign Awards and its curation thesis. Mentioned for his framing of talent as an illegibility problem and the vouching-based network model. Quote: "it's actually very illegible to see who is the right person for the right role." [00:04:47] He has "been doing this kind of thing for now for five, 10 years across Evondeck and now A6NZ." [00:14:13]

Bella Nazari

Leads growth for Cosign; previously (and still) on a16z's Speedrun team doing talent. Mentioned for the "full stack talent partner" philosophy and community-building approach. Quote: "we think a lot about how to... be the most like full stack talent person, talent partner, recruiter. And for us, that meant growing a lot of talent networks." [00:02:27]

Sophia Dew

Host of MTS and participant in the awards (a Rising Stars nominee). Mentioned as the interviewer and a data point on founder pain: "I'm constantly talking to a lot of founders every day here on MTS... they're struggling to find great talent." [00:04:02]

Theo Jaffe

Named by Nazari as a nominee on the Rising Stars list. Quote: "We do also have a lovely Theo Jaffe here. The Theo Jaffe. Yes, yes." [00:09:06] Context suggests he is a recognized new-media figure.

Sofia Puccini

Also named on the Rising Stars list. Quote: "The Sofia Puccini. Yes, yes. Oh, that's pretty cool." [00:09:06]

Brent

Referenced by Dew as a rival in Rising Stars voting. Quote: "Look, I don't need a win. I just need to beat Brent." [00:09:55] Mentioned in passing only.

5. Operating Insights

Build Community Around Emerging Role Archetypes, Not Industries

Nazari's team organizes events and lists around roles that are rapidly changing, such as the forward deployed engineer (FDE), growth leaders, and talent leaders, rather than around company types. Quote: "creating these communities around certain pieces of leadership... what are the broad-based themes of where we're seeing like roles go, work go, you know, companies go. And then putting those people in a room." [00:12:30] The tactic: identify a job title becoming load-bearing before it's well-defined, and convene its practitioners.

Use Peer and Senior-Function Nominations as the Quality Filter

Booth highlights that the most credible nominations come from within a function: "the ones I think are seeing nominations coming in from their peers, coming in from senior leaders in the functions." [00:10:07] When building any reputation system, weight vouches from domain-relevant seniors over raw vote counts.

Run a Public Review Loop to Fix Coverage Gaps Before Finalizing

Rather than closing the list internally, they asked invitees to audit it: "We've been asking people who are invited, like, review the list. Is there someone missing?" [00:10:55] Crowdsourcing the omission check improves list quality and turns recipients into promoters.

Seed the Product With a Time-Boxed Social Moment

Nazari notes the timing choice: the awards run "later in the year" when "people are grateful for people this time of year," to prompt the first use of the cosign button. [00:07:32] Booth echoes: "Coming on Thanksgiving, right? Who are you thankful for?" [00:08:29] Tying a novel action to an existing cultural ritual lowers the activation barrier for a new behavior.

Curate After the Votes: Judges Select 5 to 10 Per Category

The funnel is crowd nomination, then expert judging, then a small invite-only pool: "we're going to go back through this with our judges. We're going to work through and find five to ten people across each of the categories and invite." [00:10:55] This hybrid keeps scale from the crowd and taste from the judges.

6. Overlooked Insights

The Agent Endpoint Is the Real Product Wedge

Buried in a short passage near the end is the most strategically important fact: users are already pointing autonomous agents at Cosign and instructing them to monitor "anytime somebody that I'm connected to creates a new list, see who's on the list." [00:14:59] That is a reputation graph being consumed programmatically, with agents re-ranking and re-surfacing people continuously. If agents become the primary way people discover talent and opportunities, whoever owns the highest-trust, machine-readable vouching data (with explicit human attestations behind it) has a durable advantage over static profile databases. The awards then look less like an event and more like a bulk labeling exercise: 49 awards across hundreds of nominations produce structured lists that agents can query.

"Talent Mafia" Framing Reveals the Actual Business Model Logic

Booth's offhand phrase, "get on the ground floor of the next great talent mafia," paired with the Rising Stars criterion, "I want to be in first on the precede round of this person," [00:09:28] reveals that Cosign doubles as an early-signal sourcing system for investors and recruiters. The people nominated as rising stars are being identified before they are widely known, and the vouchers who nominate them are themselves revealed as high-taste connectors. A graph of who vouches for whom is simultaneously a talent pipeline, a deal-flow source, and a taste-ranking of the vouchers. This is a quiet but significant asymmetry for a firm positioned at the center of it.