robert

434 posts

robert banner
robert

robert

@robechun

Building @Charcoal_HQ prev - explorer @southpkcommons - AI lead @FrontHQ

San Francisco, CA Katılım Aralık 2016
310 Takip Edilen192 Takipçiler
robert
robert@robechun·
many founders warned us about delve... apparently some companies got fk'ed during raise / acquisition because delve did a bad job with audit
Ryan@ohryansbelt

Delve, a YC-backed compliance startup that raised $32 million, has been accused of systematically faking SOC 2, ISO 27001, HIPAA, and GDPR compliance reports for hundreds of clients. According to a detailed Substack investigation by DeepDelver, a leaked Google spreadsheet containing links to hundreds of confidential draft audit reports revealed that Delve generates auditor conclusions before any auditor reviews evidence, uses the same template across 99.8% of reports, and relies on Indian certification mills operating through empty US shells instead of the "US-based CPA firms" they advertise. Here's the breakdown: > 493 out of 494 leaked SOC 2 reports allegedly contain identical boilerplate text, including the same grammatical errors and nonsensical sentences, with only a company name, logo, org chart, and signature swapped in > Auditor conclusions and test procedures are reportedly pre-written in draft reports before clients even provide their company description, which would violate AICPA independence rules requiring auditors to independently design tests and form conclusions > All 259 Type II reports claim zero security incidents, zero personnel changes, zero customer terminations, and zero cyber incidents during the observation period, with identical "unable to test" conclusions across every client > Delve's "US-based auditors" are actually Accorp and Gradient, described as Indian certification mills operating through US shell entities. 99%+ of clients reportedly went through one of these two firms over the past 6 months > The platform allegedly publishes fully populated trust pages claiming vulnerability scanning, pentesting, and data recovery simulations before any compliance work has been done > Delve pre-fabricates board meeting minutes, risk assessments, security incident simulations, and employee evidence that clients can adopt with a single click, according to the author > Most "integrations" are just containers for manual screenshots with no actual API connections. The author describes the platform as a "SOC 2 template pack with a thin SaaS wrapper" > When the leak was exposed, CEO Karun Kaushik emailed clients calling the allegations "falsified claims" from an "AI-generated email" and stated no sensitive data was accessed, while the reports themselves contained private signatures and confidential architecture diagrams > Companies relying on these reports could face criminal liability under HIPAA and fines up to 4% of global revenue under GDPR for compliance violations they believed were resolved > When clients threaten to leave, Delve reportedly pairs them with an external vCISO for manual off-platform work, which the author argues proves their own platform can't deliver real compliance > Delve's sales price dropped from $15,000 to $6,000 with ISO 27001 and a penetration test thrown in when a client mentioned considering a competitor

English
0
0
4
223
Sam Blond
Sam Blond@samdblond·
$100,000 cash prize founder poker tournament. No entry fee. Announcing The Monaco Invitational, presented by @MonacoGTM . This will be the largest founder poker tournament (and party) of all time. In addition to giving away $100,000 to the winners in the free to play tournament, we'll have celebrity guests, an incredible food and beverage program, and the coolest venue in San Francisco. This will be the event that sets the gold standard for startup events. Invitations are reserved for Monaco customers, anyone who refers a Monaco customer, and friends of the firm. Know a founder who should be there? Tag them in the comments. Check out more details in the invitation on tournament rules and eligibility.
Sam Blond tweet media
English
59
18
253
68K
Aida Baradari
Aida Baradari@aidaxbaradari·
Today, we're introducing Spectre I, the first smart device to stop unwanted audio recordings. We live in a world of always-on listening devices. Smart devices and AI dominate our world in business and private conversations. With Deveillance, you will @be_inaudible.
English
1.1K
5K
42.5K
4.4M
robert
robert@robechun·
@rtwlz how do you keep finding these things
English
0
0
0
460
robert
robert@robechun·
@grinich 🙋🏻‍♂️ would love one!
English
0
0
0
53
Michael Grinich
Michael Grinich@grinich·
I wrote the book I always wanted. 📖 A complete, practical guide to enterprise readiness — how to go from “users love us” to “enterprises buy from us.” 275 pages. 18 chapters. The missing manual for founders going upmarket. Limited print run. Want a copy?
Michael Grinich tweet mediaMichael Grinich tweet mediaMichael Grinich tweet mediaMichael Grinich tweet media
English
80
6
192
11.3K
robert
robert@robechun·
@paradite_ I think it makes sense for them to see demand/appetite & get a feel for how they can get better margins
English
0
0
0
171
Zhu Liang
Zhu Liang@paradite_·
Honestly I have no idea why Anthropic is releasing a faster Opus at 6x price. Maybe it's just an experiment, but in the current competitive environment against Codex, it is a move that is gonna hurt public perception.
eric provencher@pvncher

I really appreciate that OpenAI is doing everything they can to make codex faster at the same pricing and rate limits, while Anthropic is raising the price by 6x, and locking this out of your max plan, for the extra speed.

English
26
2
116
16.6K
robert
robert@robechun·
cerebras for all inference please, i want claude code to be instantaneous. thanks
English
0
0
1
142
robert
robert@robechun·
@SherryYanJiang I’m using it! I’m equipping it with a bunch of sub-agent MCPs. I think their personality + iMessage integration is bis rn
English
0
0
1
198
Sherry Jiang
Sherry Jiang@SherryYanJiang·
what happened to the poke.com hype?? am i the only one that is still using it?? (we all too cowork and claude code pilled now?)
English
73
5
245
52.9K
robert
robert@robechun·
Either @cluely is truly going downhill or @im_roy_lee got fleeced? these kids came in, got paid $$$, gained clout, travelled a bunch & then are now peacing out using said gained clout for their next gig?
English
0
0
0
173
robert
robert@robechun·
CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE CLAUDE CODE
Français
0
0
1
64
robert
robert@robechun·
A couple assumptions I think you're making (even with the presumption that you mean 2026+, not just current state of AI coding today): 1. Even bigger, much more complex software can be created on the fly very quickly. 2. The bottleneck of creating such software will be tps. 3. Same input -> fundamentally same output, every time 4. Sub-systems / packages created by main system (agent) will be accurately spawned off by the main agent accurately and reliably, every time. 5. Predicability is no longer a first-class-citizen etc. E.g. take point #2: I posit there is a large set of enterprise software out there that a vanilla coding agent cannot conjure up (however fast and powerful it may be!)—because there are tons of tacit workflows and knowledge never exposed to the agent. You can't JIT something from information that is ~mostly/completely opaque to the agent. (I like generative apps tho! I think it makes a lot of sense in areas like personalization - but again, I don't think the generalized statement is accurate)
English
1
0
1
22
Sherwood
Sherwood@shcallaway·
Sharing an idea that has been rolling around in my head... The End of Reusable Software When something is free, you don't reuse a single instance over and over again. You get a new instance each time. Example: plastic water bottles. In 2026, code is effectively free. Why would I fork a project or install a library when Claude can implement exactly the functionality I need from scratch? Taking things a step further: if software is truly free, what's the point in creating reusable programs? Why not just write one-off programs for every scenario? AI agents already do this, e.g. when they use Python sandboxes to do math or when they create one-off bash scripts to complete tasks. Side-note: I need a good name for this. Current ideas: - Just-in-Time Programming - Generative Apps - Runtime Compilation There are arguments for re-using software, but right now I can't think of any good ones. Here are two: 1- Re-using software saves tokens. I guess if you're creating something that involves a lot of code, this could matter... But mostly I don't think this argument holds up. Tokens are pretty cheap. Electricity is pretty cheap (for now). Maybe re-using software is the "green" thing to do... 2- Existing software may be more reliable or secure than new software. In many cases, existing software has been battle-hardened, pen-tested, certified, scanned for vulnerabilities, etc. On the other hand, using existing software also exposes you to supply chain attacks... The End of Reusable Software raises a lot of big questions, such as: - What is the future of OSS? - What is the future of package registries like NPM? - What is the future of SaaS? - Where will store/save all this new software? - Is there any point in saving it? - How do we do "observability" when every program is ephemeral? - How do compute & storage primitives need to change to support this new paradigm? - How does hardware need to change to support this new paradigm? - Will I have a job next year? Would love to hear what people think about this idea. Maybe I'll write a blog post...
English
30
2
50
14.7K
robert
robert@robechun·
@vasuman Have something for you; DM'ing!
English
0
0
0
11
vas
vas@vasuman·
Biggest agents advice I can give: The space moves so fast that if you marry your agent to any tech stack, there's a chance it's obsolete in 30 days. Build modular. Not just swapping API endpoints, your entire framework needs to be editable with continuous evals and regression tests so every change is measured and reversible. Memory, orchestration, tool-calling, all of it is being productized. If you spent 2 months building custom memory architecture from the ground up, you probably wasted your time. Your time is better spent building on top of productized services. Modular = swap in the new industry standard on demand (or make the migration as painless as possible, because nothing is ever that easy in software). TLDR: Staying on top of the best frameworks is a full-time job. That's why I have a CTO @eyad_khrais who's barely on X.
English
33
17
477
62.4K
robert
robert@robechun·
As 2025 winds down I thought I'd share one life hack for paying 0 taxes on capital gains in California: trade options and be negative realized gains for the year
English
0
0
0
53
robert
robert@robechun·
so it turns out 2025 wasn't the year of agents but the year of acquisitions
English
0
0
0
39