Akash Samant

making insurance easier @ Coverflow | formerly @robustHQ

[Akash Samant @AkashSamant4]

@arshdilbagi lets go! big launch!

[Arsh Shah Dilbagi @arshdilbagi]

Introducing Adaline 2.0 - The Agent Self-Improvement Layer

Adaline turns Traces into Behaviors,
Behaviors surface Issues,
Issues become auto-generated Evals + Data,
Adaline then generates new agent candidates and tests them.

You review the winners and ship!

[Arsh Shah Dilbagi @arshdilbagi]

The hard part about LLM failures is that their outputs rarely look like failures.

The demo “works.”
The output sounds coherent.
The user actively uses the product.
And your dashboard looks normal.

Meanwhile, the system can be wrong, unsafe, or quietly driving up token spend. And you won’t notice until the damage adds up.

Prompts often serve as business logic (policies, safety, and product context). But many teams ship them without the basics, such as versioning, reviewable changes, end-to-end traces, and eval gates.

In production, it doesn’t crash. It degrades via wrong answers, policy misses, and surprise spending.
No crash. No error. No alert.

I cover this exact issue in my @Stanford CS 224G guest lecture on AI Observability and Evaluations.

Here are the core ideas:
• If you only log the final output, you’re guessing. Full traces show where it broke.
• Evals are feedback loops. Use clear pass/fail criteria tied to outcomes.
• Run evals continuously on production traces and don’t wait for support tickets.

The moat isn’t prompt cleverness. It’s a measured improvement.

Full lecture + blog below 👇

[Akash Samant @AkashSamant4]

Has been amazing to watch the crazy growth!!! @hwchase17

[Akash Samant @AkashSamant4]

@hungrysohan adorable

[Akash Samant @AkashSamant4]

@hthieblot hello! this seems crazy. would love to be there

[Derrick Li @TheDerrickLi]

Huge congrats to @FastowMatthew, @AkashSamant4 and the entire Coverflow team 🚀

Brokers waste thousands of hours each year on repetitive, manual tasks like reviewing policy docs and juggling spreadsheets. Coverflow’s platform automates all of that.

https://t.co/TzPxibNNk0

[Akash Samant @AkashSamant4]

Data + Evals = Magic.

[Arsh Shah Dilbagi @arshdilbagi]

Most AI products fail in the first month. Not bad AI. Bad prompts.

Teams at Discord, McKinsey, Salesforce, DoorDash, Reforge, and over 100K+ developers using us know why:

Teams wing their prompts – test on 5 examples, ship to millions, pray it works.

Today changes everything.

[Arsh Shah Dilbagi @arshdilbagi]

Most AI products fail in the first month. Not bad AI. Bad prompts.

Teams at Discord, McKinsey, Salesforce, DoorDash, Reforge, and over 100K+ developers using us know why:

Teams wing their prompts – test on 5 examples, ship to millions, pray it works.

Today changes everything.

[Hubert Thieblot @hthieblot]

If a founder tells you their main goal is to retire their parents

Invest in them immediately.

[Akash Samant @AkashSamant4]

finally, a solution for taking zoom calls in the shower

[ₕₐₘₚₜₒₙ @hamptonism]

yea, we’re cooked.

[Akash Samant @AkashSamant4]

@AviSchiffmann EU rights charter *

[Akash Samant @AkashSamant4]

can’t wait to go in for my monthly oil change 10 years from now

[Bryan Johnson @bryan_johnson]

I am no longer injecting my son's blood.

I've upgraded to something else: total plasma exchange.

Steps:
1. Take out all blood from body
2. Separate plasma from blood
3. Replace plasma with 5% albumin & IVIG

Here's my bag of plasma. Who wants it?
🧵

[Akash Samant @AkashSamant4]

@AviSchiffmann That would be great for our margins

[Akash Samant @AkashSamant4]

@iamgingertrash nice

[Akash Samant @AkashSamant4]

@calvinchen let’s go!

[Akash Samant @AkashSamant4]

Dune getting snubbed at the oscar’s is so surprising, but also not at all

[Free_Ross @Free_Ross]

FREEDOM!!!!

[Akash Samant @AkashSamant4]

LLMs are not a silver bullet. The way you solve a problem matters way more than the tools you use to solve it

[Hamel Husain @HamelHusain]

New post re: Devin (the AI SWE). We couldn't find many reviews of people using it for real tasks, so we went MKBHD mode and put Devin through its paces.

We documented our findings here. Would love to know if others have had a different experience.

https://t.co/DDqzoAXKkl

[Hamel Husain @HamelHusain]

New post re: Devin (the AI SWE). We couldn't find many reviews of people using it for real tasks, so we went MKBHD mode and put Devin through its paces.

We documented our findings here. Would love to know if others have had a different experience.

https://t.co/DDqzoAXKkl