Getting Ready for 2.0! Neuroscientist. ex-CTO, Founder @madstreetden @vue_ai Startup & Life adventures with @LadyAshBorg
San Francisco, CA
Joined September 2016
- Tweets172
- Following122
- Followers2.1K
- Likes269
Before vibe-coding, since it took time to solve the problem, the concepts got reinforced because we stayed longer inside the problem space. We aren't reinforcing our own memories. Giving away our evals to the frontier overlords here 🙃
Facing something very similar to @LadyAshBorg while writing code ... Correction, watching claude and codex write code. I go deep into the problem with them and solve it and then completely blank out 2 days later, when trying to explain it.
We talk so much about attention loss with all the context switching we're doing between our agents, assistants and endless AI minions. We rarely speak about memory and what it does to our ability to recollect stuff.
I've started writing again, to counter the 12-14 hours of AI use!! Letting the AI do the plan and roadmap but use the writing to orient myself here and now from that plan, doing some simple connections between thoughts, artifacts etc.
Between context loss and the chronological layout of content in these AI things ... it is becoming utterly hard to make things visible and recollecting them using the AI interfaces themselves.
I've been speaking about the AI SDLC process to multiple startups for a while. There is really just 1 rule at the end of the day. There is no one size fits all for AI DLC.
Each company should build their company brain to reflect the way they work. Its no wonder that Atlassian's version reflects what's under their hood and what products they build.
Today we're releasing our AI SDLC transformation playbook, where we share best practices to deploy AI at enterprise scale
Engineering leaders frequently ask how Atlassian thinks about our AI-native SDLC. This playbook is our answer, informed by our own transformation and learnings from top engineering organizations we work with
Come find us at the AI SDLC booth at Team '26 EU this week for a physical copy (and merch) atlassian.com/blog/ai-at-wor…
Anand Chandrasekaran retweeted
Outcome pricing only works when you can actually measure the result and the vendor controls enough of the process to deliver it. This is why companies like Palantir and Sierra can charge this way. Their products take over the whole human workflow, not just the AI model inside it. Most large companies cannot buy software this way yet, because the final result depends on their own data, their own permissions and their own ways of working. Until a vendor can promise the entire loop from start to finish, the fairest deal sits somewhere in the middle. You charge for each task completed and you commit to a result you can show.
Ali makes great points.
In addition, I think one of the key reasons for lagging enterprise AI adoption is that F500 enterprises don’t buy tokens. They buy business outcomes.
Palantir and Sierra are arguably the fastest growing enterprise AI applications companies (outside of the labs). Both sell business outcomes and price on outcomes.
If you’re selling tokens to F500 enterprises, you’re DOA.
Anand Chandrasekaran retweeted
Jev will probably go down in history as the most thoroughly spanked startup ever — a brutal reminder of how few ways are are to create a competitive moat in software these days.
(Hint: if you don’t have proprietary data that clearly distinguishes your offering then watch out.)
1. Launch a service that popularises a novel and valuable way to use LLMs
2. Watch literally everyone clone it almost instantly, in many cases beating it
Cloudflare's Clef beat Jev in just two weeks
How: "frozen" Qwen does one prefill pass, a tiny schema head scores every answer option in parallel, with no text generated at all.
That's why it's 4x faster than Jev at 2x the accuracy on some benchmarks.
And it's open source!
Link to HF:
huggingface.co/Cloudflare/cl…
The speed isn't the story. It tells you which layer just became cheap to copy.
The sheer number of Jevalikes/Jev wannabes that have shown up in the last week is fascinating. That an Amazon or its labs can respond this quickly is pretty wild. Also because of what it says about how quickly the system now corrects for perceived inefficiencies.
The moment someone demonstrates a plausible new abstraction, hundreds of people are already studying it, decomposing why it works and trying to exploit whatever opportunity it has revealed.
We're seeing almost instantaneous feedback loop: inefficiency appears > someone finds a plausible solution > solution becomes visible > capital and builders swarm it > dozens of variants appear > the market discovers which parts were actually valuable.
Some days it feels like the entire system is working in unison, iterating off of each other. It is truly a wild wild time to be building.
The interesting thing about Jev or any other leaps we're making today may ultimately be less about each of them, and more the speed with which the ecosystem is assimilating and responding to what they reveal.
The sheer number of Jevalikes/Jev wannabes that have shown up in the last week is fascinating. That an Amazon or its labs can respond this quickly is pretty wild. Also because of what it says about how quickly the system now corrects for perceived inefficiencies.
The moment someone demonstrates a plausible new abstraction, hundreds of people are already studying it, decomposing why it works and trying to exploit whatever opportunity it has revealed.
We're seeing almost instantaneous feedback loop: inefficiency appears > someone finds a plausible solution > solution becomes visible > capital and builders swarm it > dozens of variants appear > the market discovers which parts were actually valuable.
Some days it feels like the entire system is working in unison, iterating off of each other. It is truly a wild wild time to be building.
The interesting thing about Jev or any other leaps we're making today may ultimately be less about each of them, and more the speed with which the ecosystem is assimilating and responding to what they reveal.
Anand Chandrasekaran retweeted
Bot detection & human verification will be one of the most urgent demands for businesses over the coming years. Agent swarms will suffocate every website and form; small companies and government websites are most vulnerable.
There is a huge gap in the market for this right now. When we looked at what offerings were in the market to use at X, there was not a single company that brought together all the latest technologies so we had to do it all in-house.
Anand Chandrasekaran retweeted
Absolutely thrilled to announce M2P-MSD upcoming deal. The 'no stress, not distressed' coverage by The Hindu :D thehindubusinessline.com/inf…
This is what happens when you get founders at a table drinking more coffee than they should have.
Jokes apart - Not used to crisis PR. But like the fabulous Billie Jean reminder: Pressure is a privilege. Responsible journalism over clickbait. So here's our actual story at MSD.
The most repeated question: Why?
Very simple answer: We finished our vision for building the generalizable AI platform we began with. Continue to power some of the world's largest, best enterprises. Compete with Palantir, Databricks, UiPath and more. At some point in every founder's journey, you know what you want and what's the next right direction. We've never wanted an IPO story, the M&A was always the exit playbook.
The Process: Declined a series D at the start of 2024 and got the board together towards an M&A at the start of 2024. Prepared for 6 months for an M&A and launched the process with one of the best bankers in the US. Courted some of the absolute winners in the US market, moved into offer stage, when M2P happened completely from left field. To be able to land MSD with another Chennai star, an IPO-bound story is the stuff of dreams. Completely unexpected. Heavy overlap of customers. And a vision to scale this AI first story across adjacent industries together globally.
The Platform: Vue.ai has been a horizontal platform serving industries from all backgrounds. Over the years, it has evolved from delivering SaaS products to becoming a unified data and AI orchestration platform for multiple industries. The platform has grown from strength to strength serving up more and more capabilities to our customers requiring fewer and fewer standalone products in their tech stack mix. From AI based applications to building your own models and creating data transformation workflows, industry giants use these to power their logistics, banking, finance, healthcare use cases. We'll be releasing a huge AI agent library, scaling our data platform play and doubling down on our marketplace in the coming quarters. Our customers are just as excited with the news and the might of the joint scale, its go time!
The People: Anand, Costa and I are going with our fabulous teams into the org and expect to double down on M2P's inspiring growth story. Over 25+ staff have 8+ years of sweat and blood building this story. Another 40 with over 5+ years. Many more with fewer years and going on this journey longer.
We've had so many mentors, investors, champions by our side during this period. Building a company like MSD at a time when no one even knew what AI was ... required risk takers of a whole other kind. You know who you are. Thank you for everything. All the highs and lows and the invaluable lessons. To all the friends/founders who were there for all the midnight panic calls and crying. It truly takes a village. Thank you!
We'll be sharing a ton more in the coming weeks as the deal actually closes, legally. If you are a founder, you'll hear from us at the SaaSBoomi annual event. If you are looking to get hired, you must talk to the team that's getting acquired
@onlymadhoo @prabranga @SirBlueBorg will be sharing much more ... once this is all actually done
Anand Chandrasekaran retweeted
Here's @SindhuHarih in The Hindu :)
A full scale tweet story coming up as promised from yesterday. Over the moon to be announcing what is coming up. And how incredibly exciting it has been for our customers and teams alike to make this choice over the rest.
Read The Hindu for all the facts.
The ☕️ on the M2P-MSD deal
Founders call it a “perfect synergy story” and say deal is valued in “multiples” of what’s out there
@MadStreetDen had US suitors and funding round but decided on @m2pfintech for cultural fit and customer overlap
thehindubusinessline.com/inf…
Anand Chandrasekaran retweeted
Two #Chennai #startups come together in what they call a "perfect synergy" putting culture first...
"@m2pfintech is in talks to acquire @MadStreetDen ...The deal is valued way higher & in multiples of what is being quoted" founders say
More details:
thehindubusinessline.com/inf…
Anand Chandrasekaran retweeted
The ☕️ on the M2P-MSD deal
Founders call it a “perfect synergy story” and say deal is valued in “multiples” of what’s out there
@MadStreetDen had US suitors and funding round but decided on @m2pfintech for cultural fit and customer overlap
thehindubusinessline.com/inf…
Anand Chandrasekaran retweeted
Chennai-Chennai
One more win for Chennai
Super excited to share …
Stay tuned for the full facts from us shortly! @onlymadhoo
Anand Chandrasekaran retweeted
While the rumour mills are abuzz. Wait until you hear from the source than sources.
Anand Chandrasekaran retweeted
Will AI take our jobs – or make use more efficient? @MadStreetDen’s @LadyAshBorg chats with Peak XV’s @smdcmc about the state of Enterprise AI, her experience as an early player in the space, and why she feels everybody needs to become ‘AI native’. youtube.com/watch?v=AXdZ3T4H…
Anand Chandrasekaran retweeted
Fedex invests in MSD to accelerate Enterprise use of AI.
Very excited to welcome Fedex as an innovation partner, a customer, and an investor on this journey to make large enterprises AI-Native.
This is a testament to our rapidly growing Enterprise presence in Fortune 500 list
FedEx invests in Mad Street Den to accelerate enterprise use of AI.
@FedEx will use advanced technology & data-driven insights to create smarter supply chains with differentiated offerings to deliver greater value & experience for customers.
Read more: tinyurl.com/madstreetden
