Clad is an AI support engineer aimed at technical teams whose engineers keep getting pulled off product work to answer customer questions. It pulls messy threads from Slack, Discord, Teams, and email into a single line of investigation, then connects to billing, usage, logs, and code to figure out what actually happened. The output is an evidence-backed handoff, meaning a root cause, supporting evidence, and a ready-to-file ticket rather than another vague escalation. Clad triages, investigates, and resolves support issues end-to-end, and it's backed by Y Combinator .
The launch matters because support has become the hidden tax on fast-moving engineering orgs. One customer question can send a senior engineer through six systems to reconstruct context that has been reconstructed before, and Clad is positioned as the investigation layer that sits between support and engineering so that work compounds instead of repeating. Every resolved issue becomes memory the agent can reuse on the next ticket, which is what makes the "resolves itself" framing more than a tagline. The company also claims teams can wire it into their existing tools and tickets in about 30 minutes, which lowers the bar for trying it during a busy quarter.
Clad Labs is a Y Combinator F25 company founded in 2025 by Kevin Le and Richard Wang . Richard Wang is the co-founder and CEO and is posting the launch on X as @zzzbuggg. If you run a technical support function or you are an engineering leader tired of watching your best debuggers turn into human search engines, this is a launch worth a closer look.
Comments (15)
Sign in to join the discussion.
N
Nadia OkaforSep 3, 2026
ok wait, 'support that resolves itself' is a wild flex given half the tickets I file are begging a human to please just read the previous message.
P
pacecutsSep 3, 2026
the cut at 0:14 where the ticket auto-closes hits, but the voiceover is 3db too loud over the demo audio. otherwise a clean edit.
M
Mirko VasiljevicSep 3, 2026
hot take: every 'AI support' launch in the last year had this exact same tweet structure. the thread emoji is doing heavy lifting.
A
Amina BekeleSep 3, 2026
hey, would love to chat for a piece on AI in customer support. what's the wedge you're leading with vs the incumbents? happy to DM.
T
Tulasi RamaswamySep 3, 2026
how does it handle the tickets where the user just types 'it's broken' and logs off forever. asking as someone who has been that user.
B
Benoit LamSep 3, 2026
the resolve-itself framing is smart positioning. most competitors still market 'deflection' which sounds like something you do to a customer, not for them.
R
Reeta NguyenSep 3, 2026
onboarding question: how many clicks from signup to first resolved ticket? if it's more than three I'm going back to a shared inbox and my will to live.
S
sam d.Sep 3, 2026
demo videos always show the ticket that was going to resolve anyway. show me the angry enterprise customer on ticket #47 about SSO.
J
Joaquin PereiraSep 3, 2026
anyone hiring forward deployed for AI support tools? I have a candidate who ran ops at a fintech and can write SQL in her sleep.
H
Harun Y.Sep 3, 2026
been building something adjacent for 3 weeks, was going to launch friday. now I'm going to sit in a dark room and rewrite my landing page.
L
Lena KowalczykSep 3, 2026
the thumbnail is doing that thing where it's a screenshot with one arrow. worked for every good launch this month so I can't even hate.
T
Tomo I.Sep 3, 2026
this is probably the last support product launched without a full agent loop under the hood. the next wave won't even have a ticket UI.
K
kevSep 3, 2026
resolved tickets on-chain when. genuinely think there's a proof-of-resolution angle here nobody has explored.
P
Priya S.Sep 3, 2026
curious about your reply-to-view ratio on this thread. the first tweet looks like it's pulling well but the follow-ups usually cliff around tweet 4.
M
Maxim GreySep 3, 2026
the best support ticket is the one that never gets filed. build for that and you win, chase deflection metrics and you're building a smarter macro.
Work With Us
Book an intro call to explore our suite of services. We might have a waitlist.
Clad Lab
Support that resolves itself.
Support that resolves itself.
Tags
TLVC Rating
Community Rating
No ratings yet
to rate this launch.
Clad is an AI support engineer aimed at technical teams whose engineers keep getting pulled off product work to answer customer questions. It pulls messy threads from Slack, Discord, Teams, and email into a single line of investigation, then connects to billing, usage, logs, and code to figure out what actually happened. The output is an evidence-backed handoff, meaning a root cause, supporting evidence, and a ready-to-file ticket rather than another vague escalation. Clad triages, investigates, and resolves support issues end-to-end, and it's backed by Y Combinator . The launch matters because support has become the hidden tax on fast-moving engineering orgs. One customer question can send a senior engineer through six systems to reconstruct context that has been reconstructed before, and Clad is positioned as the investigation layer that sits between support and engineering so that work compounds instead of repeating. Every resolved issue becomes memory the agent can reuse on the next ticket, which is what makes the "resolves itself" framing more than a tagline. The company also claims teams can wire it into their existing tools and tickets in about 30 minutes, which lowers the bar for trying it during a busy quarter. Clad Labs is a Y Combinator F25 company founded in 2025 by Kevin Le and Richard Wang . Richard Wang is the co-founder and CEO and is posting the launch on X as @zzzbuggg. If you run a technical support function or you are an engineering leader tired of watching your best debuggers turn into human search engines, this is a launch worth a closer look.
Comments (15)
ok wait, 'support that resolves itself' is a wild flex given half the tickets I file are begging a human to please just read the previous message.
the cut at 0:14 where the ticket auto-closes hits, but the voiceover is 3db too loud over the demo audio. otherwise a clean edit.
hot take: every 'AI support' launch in the last year had this exact same tweet structure. the thread emoji is doing heavy lifting.
hey, would love to chat for a piece on AI in customer support. what's the wedge you're leading with vs the incumbents? happy to DM.
how does it handle the tickets where the user just types 'it's broken' and logs off forever. asking as someone who has been that user.
the resolve-itself framing is smart positioning. most competitors still market 'deflection' which sounds like something you do to a customer, not for them.
onboarding question: how many clicks from signup to first resolved ticket? if it's more than three I'm going back to a shared inbox and my will to live.
demo videos always show the ticket that was going to resolve anyway. show me the angry enterprise customer on ticket #47 about SSO.
anyone hiring forward deployed for AI support tools? I have a candidate who ran ops at a fintech and can write SQL in her sleep.
been building something adjacent for 3 weeks, was going to launch friday. now I'm going to sit in a dark room and rewrite my landing page.
the thumbnail is doing that thing where it's a screenshot with one arrow. worked for every good launch this month so I can't even hate.
this is probably the last support product launched without a full agent loop under the hood. the next wave won't even have a ticket UI.
resolved tickets on-chain when. genuinely think there's a proof-of-resolution angle here nobody has explored.
curious about your reply-to-view ratio on this thread. the first tweet looks like it's pulling well but the follow-ups usually cliff around tweet 4.
the best support ticket is the one that never gets filed. build for that and you win, chase deflection metrics and you're building a smarter macro.