Ramesh May @rameshmay

@dharmesh Congrats on the agent builder launch. At @GoPitCrew, we have created a similar builder harness but for financial services, with evidence trail while maintaining ground truth to firms policies.

@sumanthvepa How about for design ? Do you use Claude design at all ? Or rely on the old faithful Figma ?

To be honest, for the vast majority of the work I'm doing Sonnet 5 and Gpt 5.6 Terra are more than enough. Even haiku and luna are fine for much of the work. I feel inadequate. Less of a man for not needing the biggest baddest models. :-)

What's the org and culture for an ai company ?

@

57 years ago

0 0 0 0 0

One pattern I find useful for working with LLMs is a nice long ramble session. Sometimes the LLM needs more bits to understand what you're trying to achieve, but you're too lazy to type them. In these cases I like to lean back, switch to /voice and just ramble for like 10 minutes, total mess, anything goes, full stream of consciousness. Sometimes I declare it up top, something like "switching to speech recognition sorry for any typos...". Sometimes I turn it into a small interview of a few turns. But I find that the LLMs are somehow very good at reconstructing long incoherent rambles and often their echo of your own tangle of thoughts comes out quite a bit cleaner than what you started with. The result is that you improve the mind meld and have to correct things less from that point on.

No one likes to say it, "we ship our org chart." But it's true! If you truly understand the nature of how you build & what you ship, your culture is your product. Here is what we understand deeply at @GoPitCrew : AI Agents are 2 hops from the end user! Agents we use: Internal harness X Claude Code Us: product managers & engineers who read traces and recover "No, retry with X" and it recovers Human in the loop absorbs every error Reliability bar: 1X Agents our Agents build: PitCrew harness in buyers' hands Builders with no view into our harness internals Our errors silently become their defaults Our assumptions looks like a bug in their agent Reliability bar: 10X Agents our Agents ship Service bots, workflow agents End users who leave on the first failure No fallback, no patience, no debugging Our gap arrives amplified for every customer Reliability bar: 100X Everything we do at work everyday is keeping in mind those hops and its impact on product reliability. We built a culture & org around this understanding of agents. And that's what we ship! #Harness #Agents #EnterpriseSoftware

@NYCMayor NYC named their initiative: Pitcrew :) @GoPitCrew

Models and harness going through incredible iterations as we speak.

Marc Benioff @Benioff

2 weeks ago

Got ZDR? 🚀 The currency of AI is trust. Since launching Agentforce Trust Layer with Zero Data Retention (ZDR) on June 12, 2023, we’ve drawn a hard line with our model suppliers: Your data is YOUR data — it is NOT our product or their product. 🛡️ We have never used customer data

45 72 474 553K 157

Building my 3rd company. Risk taking the only muscle we need. We can't expect to "accelerate the build-up of the risk capital universe" while worrying of fakes. Failure is only path to building an ecosystem.

Krishnan @cvkrishnan

2 weeks ago

This issue of separating the fakers from actual makers is indeed a big one in the Indian ecosystem. But the other side also exists. Genuine new deeptech startups founders have low channels of visibility or access to the funding ecosystem incl. RDIF. @neelkanthmishra

1 4 19 988 3

Incredible! Orchestration of AI in an org is now a cultural problem not technical one.

Praveen Neppalli @praveenTweets

2 weeks ago

Agentic AI adoption is on fire at @Uber, and it's changing the way we build, not just in engineering, but across the entire company. Today, 99% of our engineers use AI tools. More than 70% of pull requests are attributed to local or cloud agents. And our engineers have built

177 366 3K 1.7M 5K

@levie Great points .. actually the need for considering model mix - frontier, SLM, open weights etc applies to any stage of the product maturity. Leaving EVERYTHING to frontier models will only result in bad tokenomics

@gokulr 2 of 2 The spec becomes the governance layer. That's what we're building at @GoPitCrew, connecting user intent to agents that execute within boundaries. This is not an option but mandatory for regulators to sign off.

@gokulr 1 of 2 Yup, the product spec is the contract between human intent and agent execution. And this becomes even more critical in regulated industries. It's not just what to build, it's the guardrails and compliance requirements the agent must follow every time it runs

@sv_cloud_expert @rishikulkarni @TweetSamG Guardrail compiler is a really good framing. The evidence requirement and adhering to guardrails is the hard part, not generating the output. This is where most agent systems fall apart and even more so in regulated environments.

Agentic systems have accelerated the transition from SoR to SoW. But that transition doesn't work unless you build the foundational pillars of Intent, Goals, and Guardrails first. At @GoPitCrew, we focused on building the execution harness to generate verifiable outcomes

Rishi Kulkarni @rishikulkarni

3 weeks ago

Saas was fun! @TweetSamG , @rameshmay and I have spent over a decade building and scaling software led businesses. We operated in a broad software category called System of Record (SoR), a central repository designed to store business data (customer info, employee records, or

4 2 6 279 2

Wish this post had more visibility. What a brilliant note on lesser known but thriving luxury ecosystem in India.

Asha Jadeja Motwani @ashajadeja

4 weeks ago

For my US friends: So we took an 8 seater from Abu Road to Ahmedabad. India. 25 minutes. We were given 2 excellent pilots - experienced, professional and smart! Indian jet companies are way more generous and large hearted than American jet companies. And half the price.

keyboard_arrow_left Previous keyboard_arrow_right Next

66 36 683 195K 256

@sujoycd Conference

Thank you team AWS AI for the partnership. Automated reasoning is very critical service as AI adoption grows in enterprises especially for regulated industries like financial services. Appreciate the support.

AWS AI @AWSAI

4 weeks ago

PitCrew built AI agents for financial services that can mathematically prove their outputs are correct. Every output is verified against regulations and business policies using neurosymbolic AI. Not probably correct. Provably correct. With a full audit trail. -30 minutes

1 5 6 993 3

PitCrew built AI agents for financial services that can mathematically prove their outputs are correct. Every output is verified against regulations and business policies using neurosymbolic AI. Not probably correct. Provably correct. With a full audit trail. -30 minutes instead of 2 weeks for manual cross-referencing -10 minutes to onboard clients, down from 3–4 hours -30 seconds to clear marketing posts vs. a 3-day compliance review Your experts go back to doing expert work. Built with Automated Reasoning checks in Amazon Bedrock Guardrails. go.aws/4eICHWO

Greying out and partially hidng the cancel subscription is NOT COOL

How to frustrate customers by HIDING the cancel subscription .. can you find it ? Is this the new-age user retention strategy ? #badux