Raj Tiwari
@RajWebshar
Followers
397
Following
14K
Media
340
Statuses
5K
Passionate about building great products and great teams. Building Trenta. Previously @iitbombay, @NuanceEnt, @Apple, @healthfidelity
SF Bay Area
Joined November 2011
Security or Compliance? . There is always someone willing to sell you a security solution you do not need. Let's get you thinking the right way. A 🧵👇.
2
1
10
So much hype on AI image generation, yet nothing can generate a halfway decent passport photo I can actually use.
1
0
4
Blazing-fast image creation – using just your voice. Try Grok Imagine.
278
544
3K
If you're wondering why your AI model just dropped a PhD dissertation explaining the 5,000 lines of slop it added to your repo, understand that AI models are paid by the token.
0
0
2
When traveling, it helps to have a portable ergonomic setup (with Roost stand)
0
0
3
Been hearing a lot about @claudeai Code, so decided to give it a try. Code spinning it's wheels unable to bring up the tutorial. Finally gets it after many tries. @cursor_ai , on the other hand, has the right solution right away 🤔
0
0
0
There is no Hindi word for honesty.
I want to crowsource more of these across languages - Please share yours!. I'll start:. There is no hindi word for nuance.
0
0
0
Reminds me of my meeting with a team that had just transitioned from waterfall to sprints. I asked them about their sprints. "We do 2-week sprints!"."We do a sprint zero where we write all the requirements. Then 7-8 sprints of dev, then 2 sprints of QA".
I recently came across a team using an LLM to implement a state machine. They called it "agentic". They also told me it worked correctly most of the time. I asked them to write down the list of rules for moving from one state to another. They listed them out. It didn't take.
0
0
2
Didn't realize the VC title came with a license to profile racially.
Have met two founders in SF this month. Both fraud:. 1. Is subletting a rented apartment and showing that as revenue for his startup. 2. Is claiming Amazon and Google are clients who have signed LOIs when they have never even heard of them. What’s common among them? Both desi men.
0
0
2
I wrote a post along these lines .No, You're Not "Just a GPT Wrapper": The Rise of Vertical AI.
linkedin.com
We’ve all heard it. The dreaded term: “GPT wrapper.
0
0
0
This rings true. At Trenta, we're finding that true value comes from getting the nuances of the audit frameworks and the style of our clients right. This means wrapping LLMs in our process instead of being an LLM wrapper.
Getting AI Agents to work extremely well for complex enterprise use cases is non-trivial. If you’re building agents, your moat will directly correlate to the amount of software you have to build on top of the AI models to execute the task. The harder the problem the better.
1
0
0
No, You're Not "Just a GPT Wrapper": The Rise of Vertical AI
linkedin.com
We’ve all heard it. The dreaded term: “GPT wrapper.
0
0
1
Everyone’s calling everything a “GPT wrapper.” I think that misses the point — so I wrote about what real vertical AI looks like, and why domain depth is the moat. Curious what other founders and investors think. (link to article in next post)
2
1
2
Really impressed with @GoogleAI search AI mode. Son was stuck with homework. Google Search AI offered help without giving the answer. This is like having Khanmigo but for all homework.
0
0
7
In computing, we differentiate between IO-bound tasks (limited by input/output operations) and CPU-bound tasks (limited by processing power). I've realized that when using AI tools in my work, a similar distinction emerges:. "IO-bound" mental tasks: These involve translating your
0
0
1
On Coding With AI.One summer, Texas Instruments donated a treasure chest of equipment to a lab at my college. Buried inside were shiny development kits, intriguing software emulators for DSP chips, and massive manuals with that crisp, fresh-paper aroma. Eager and slightly naive.
0
0
1
At this point, any email that begins with "I hope this message finds you well" is 100% written by AI. This is so common that I'm surprised LLMs haven't handled this yet.
0
0
2