Reader, Son, Husband, Father and a CEO of QBrainX Inc
Orlando, FL
Joined September 2008
- Tweets3.5K
- Following1.2K
- Followers271
- Likes5.8K
Lakshman Dhasarathan retweeted
this might be my last normal tweet before biotech goes vertical
i'm not a whistleblower. I don't work at Lilly, Novo or any of the big labs
but there is a massive shift happening in human biology right now.
it's sitting in press releases and almost nobody reads them
the last 6 months alone:
- a once-daily pill roughly doubled median survival in metastatic pancreatic cancer (13.2 vs 6.6 months)
- a one-time gene editing infusion cut hereditary angioedema attacks 87%, and 62% of patients stayed attack-free and off all other prevention for 6 months
- a one-time gene edit for cholesterol dropped LDL up to 62%, and it held for over a year
- a personalized mrna vaccine beat keytruda alone in a 1,137 patient melanoma phase 3
- retatrutide hit ~28% weight loss in a phase 3
- an ai-designed drug entered phase 3
- a pill that regrows hair passed its second phase 3
- an immunotherapy that could let some rectal cancer patients skip surgery, chemo and radiation just got fda approved
- a man lived 271 days off dialysis on a pig kidney
- human cells were caught passing chromosome-sized chunks of DNA to their neighbors, and it changes how the neighbors behave
we are on the verge of an era where the human body is a software problem and the update cycle gets shorter everyday
stick around, it's going to get weird
bio/acc
Life happens and it’s not fair at times. Those that step in and fix for the humanity is made different!
@elonmusk lives with us for a reason!
Lakshman Dhasarathan retweeted
This is cool. Opensourcing detection is great. As we get AI agents operating at machine speed, we are going to need inline prevention. Interestingly we were using Koi for this an year ago and their roadmap was headed in this direction.
We welcomed the Koi team to PANW a few months ago, the combination of detection and inline prevention from XDR is what is going to be required in the longer term - detection and inline prevention.
Looking forward to launching prevention as a compliment to your efforts.
AI agents are everywhere at @Uber. It’s great to see, but the thing that keeps me up at night is how we are going to secure them. This is something that I have been thinking about for a while.
Today, our agents run 50,000+ sessions per day across thousands of endpoints. And this isn't just engineering anymore. Employees across the company use agents that read code, run commands, call internal tools, analyze data, and act on real systems.
That scale forced us to confront an important question: How do you secure agents when your security tools can't even see them?
Traditional Endpoint Detection & Response (EDR) sees the file write, but not the prompt that triggered it. It sees the network call, but not the agent's reasoning. The intent, the thing that separates malicious from benign, is invisible.
So we built Agentic Detection and Response (ADR):
• Capture the full causal chain: prompt → reasoning → tool call → outcome, across Cursor, Claude Code, Codex, and every agent our employees use.
• Triage cheaply: a fast, high-recall first pass handles the flood of benign sessions.
• Reason deeply: only suspicious events get expensive LLM analysis, enriched with source code, threat intel, and policy context.
• Red-team continuously: an offline explorer evolves hard attack variants before attackers find them.
After 10+ months in production, the results speak for themselves:
• Hundreds of credential exposures detected across 26 categories.
• Shift-left prevention blocking secrets at 97.2% precision, before they ever leave the laptop.
• Zero false positives on our enterprise benchmark, with 2-4x the F1 score of state-of-the-art baselines.
• Every attack detected on AgentDojo, the public prompt injection benchmark.
Just as valuable as the detections are the lessons from running this in production:
• The workflow is the unit of security, not the individual tool call. Attacks hide in causally-linked chains that look benign step by step.
• Credential leakage is a far more common operational issue than prompt injection.
• Approval fatigue is real: when users approve 50+ actions per session, human oversight becomes a rubber stamp.
You can't secure agents you can't observe. And nobody can solve this alone. That is why we recently joined the Open Secure AI Alliance (OSA), and why today we're taking the next step: open-sourcing ADR.
The release includes the ADR Sensor, the detection framework, and ADR-Bench, the first enterprise agentic AI security benchmark: 302 tasks derived from real production telemetry and full coverage of all 17 attack techniques across 5 tactics, so the community can rigorously evaluate their own defenses.
Code: github.com/uber/ADR
Paper: arxiv.org/pdf/2605.17380v1
The future of AI security won't be built behind closed doors. Excited to see what the community builds on it, and what we all learn together! @UberEng
Lakshman Dhasarathan retweeted
A $100K/year customer gets white-glove treatment while a $10/year customer gets a help center. @DecagonAI's bet is that AI closes that gap, and enterprise is buying it.
Co-founders Jesse Zhang and Ashwin Sreenivas sit down with a16z's Kimberly Tan and Sarah Wang to discuss what they've learned running agents inside the biggest banks, airlines, and telcos:
- Smart-vs-cheap models are a false trade-off
- Start with frontier models on new use cases, then migrate to open-source after they mature
- Fine-tuning per use case works at the application layer
- Support demand always outran supply: make it cheaper and companies buy more of it
00:00 Intro
01:07 90% of Decagon runs on open-source
05:00 The smart vs. cheap trade-off is false
09:26 Building a model factory in-house
15:07 Are the labs the last startups?
21:21 Are forward deployed engineers a trap?
28:36 The agent that builds the agent
37:02 Glass box beats black box
47:55 From help desk to AI concierge
1:14:45 Jevons paradox in customer support
@thejessezhang @AshwinSreenivas @kimberlywtan @sarahdingwang
Lakshman Dhasarathan retweeted
Before you buy another GTM tool, check whether Claude Code can already run the one you have. The answer is yes far more often than it was six months ago.
We split the GTM stack into nine layers. This is what each one is for.
→ Signal and intent. Who's in market right now. PredictLeads, Common Room, RB2B, Warmly. This is the layer we still bridge with n8n or an Apify workflow.
→ Research and web data. Study the account. Origami, Firecrawl, Exa, Perplexity, Browserbase, Apify. They read the real site copy, the live job posts, the actual profile. People skip this layer, and it's why their AI outbound reads like AI.
→ Data and enrichment. Build and verify the list. LeadMagic, Apollo, Clay turn a company name into a person with a working email. Everything downstream is capped by how good this layer is.
→ Orchestration, the brain. Decides what runs. Clay and n8n handle what fires, what waits, and what happens when a step fails.
→ Action and outbound. Get the message out. Lemlist, Instantly, Smartlead. Send caps, warmup and sender reputation all live here, which is why this is the layer I'd keep a person on.
→ CRM and system of record. Where it lands so the next person has something to pick up. HubSpot, Attio, Salesforce.
→ Infrastructure. Supabase, Pinecone, Airtable, GitHub, Vercel, for the parts you end up building once the off the shelf version stops fitting.
→ Comms and content. Slack, Notion, Google Workspace, Figma, so the output shows up where the team already works.
→ Revenue. Point Claude at your pipeline and ask what closed.
We wire this for 100+ AI companies. You don't need all 28. Start with one per layer, all pointed at the same Claude Code window, and the second one is where you'll stall, because the moment two of them can touch the same record you need a rule about which one wins, and nobody writes that rule until something overwrites a field it should have left alone.
The video is the outbound half of it, end to end, in four Claude Code prompts.
Lakshman Dhasarathan retweeted
part one of "Why Software Factories Fail" hit 250k views in under 24h - now you get to enjoy part two
Lakshman Dhasarathan retweeted
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.
AI will transform every industry, power every company, and be built by every country.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
The world needs both frontier closed models and frontier open models.
images.nvidia.com/pdf/Open-W…
Lakshman Dhasarathan retweeted
wtf is a graph.
Loops made Agent behavior programmable. Graphs make agent orgs programmable.
Dynamic Agent Org is where the graph rewrites itself while the work is happening.
Lakshman Dhasarathan retweeted
Hosted a dinner last night with a group of IT leaders of large enterprises around agent adoption in the enterprise. Some quick notes:
* Change management remains one of the biggest topics for driving workflow transformation. Still most processes need to be upgraded to modern operating models to work with agents, which is a mix of technology, data, and human process change. Lots of emphasis on getting data (structured and unstructured) into a setup that can work with agents properly.
* IT teams are finding increasing success embedding full engineers into the business functions (essentially internal FDE) that go and implement agents into the internal workflows. There’s so much technical work to be done to make agents successful, that they can accelerate months or quarters of failed experiments by having someone technical in the workflow early.
* Consensus that the tech function is becoming more important than ever. It’s clear that the business could only expect automation to affect a minority of the business before (e.g. ERP) but now it can impact all of knowledge work. This means IT is becoming a more central role to the workflows across the company.
* Workflows are cross functional, and getting agents to work cross functionally is a complicated data modeling and permissions issue. Single users don’t have access to this. Which means you need to have agentic systems take on their own roles and have their own privileges, which is non-trivial given agents can’t keep things secure on their own.
* Huge variance in budgets between coding work and the rest of knowledge work. Some companies had a $1,000 a month budget for developers, and others had much higher amounts (like $5,000) that were merely triggers to notify the team vs. block them. Far smaller budgets for non-coding work at the moment.
* More companies are building their own multimodel systems for routing workloads by task to frontier and lower cost models. Lots of energy around open weights models, but still more in experimentation instead of at scale usage (some companies can’t due to perceived Chinese issue).
* Clear sense that all enterprise software must be headless in the future. Relief that they don’t need to train employees on hundreds of different apps. However, clear frustration with the traditional vendors that don’t play extremely nice (technically or cost wise) with agents in a headless fashion. Huge warning for existing software vendors.
* Mythos or mythos level-models are finding more and more sophsiticated security risks. The chaining together of vulnerabilities is what’s novel right now, and companies are coming up with long backlogs of what they need to go patch quickly.
Even more discussed, but just a few of the hottest topics.
Lakshman Dhasarathan retweeted
This is a practical, thoughtful proposal to win cooperation. And Demis has done an admirable job, personally and professionally, to have the credibility to play this role.
Stepping back, I kind of can't believe the craziness of this moment. What's about to happen has never happened.
It's not clear the mental models we have of ourselves and the world will be relevant in the future.
Each of us is trying our best to use our pattern matching skills to predict what will be. We talk about the immediate things that we can understand like jobs and meaning making, but the scale of what's at stake, I think, is far more consequential than that.
It's possible that the speed of it all, plus the increased complexity, will feel bewildering. It's also possible that it will feel natural and intuitive.
I do hope that we've hit peak fever pitch in acrimony and division and that we'll turn a corner and become more cooperative.
Lakshman Dhasarathan retweeted
I made myself a personalized podcast player using @grok 4.5 and @perplexity_ai Computer
It finds the top topics people are discussing on tech and business podcasts, then lets me play a single stream across multiple podcasts by topic.
When you click PLAY DEEP-LINK STREAM, it plays me all the pods talking about China and open source (in the screenshot here)
it's called PODMEME (after Techmeme, which is what i modeled it off of!)
Thanks @AravSrinivas @elonmusk for the harness and model! Cost me 1,100 "credits" on Perplexity -- so $11!
Not sure what it will cost to run this every 48 hours... but, I think worth it.
I might actually output a custom MP4 file for each topic with an audio transition in the middle
Lakshman Dhasarathan retweeted
Every team is going to need a Single Brain.
-Your CRM knows the deal history.
-Gong knows the customer language.
-Slack knows what the team is actually doing.
But none of it compounds if it lives in separate tabs.
You wouldn't let your team work with no shared memory, so why are your tools still doing it?
This is the idea behind Single Brain: one shared memory layer for market intel, strategy, content, campaigns, measurement, and client context.
Lakshman Dhasarathan retweeted
We started 8090 a few years ago on the belief this motion would become widespread. We’ve been surprised by how fast it has taken hold. That said, it’s still so early in this game.
Software license revenue is $1T annually. Services and T&M consulting around those licenses is $4T annually.
All of this will merge into a new $5T super-category and whatever happens from there is anyone’s guess. But I applaud $SBUX for ripping off the bandaid - they clearly see the forest from the trees.
Starbucks spends $400 million a year on software. Yesterday they announced they're moving off IBM and Microsoft to build their own custom systems in-house.
IBM dropped 3% and Salesforce dropped 4% on the news.
And honestly this is, unequivocally, the biggest signal I've seen since OpenAI and Anthropic launched their consulting arms back in Q1. The largest companies in the world are done paying for software that half fits how they work.
We saw this coming about a year ago. Moved everything we build off Airtable and low-code tools and went fully custom. Already paying off, and it's only going to compound from here.
This is the opportunity right now.
You get all of a company's data into one system. You build out a single operating system for the entire business. You cut out bad, redundant processes. Then you layer AI on top of it, under the correct processes.
That's the core of AI consulting. Helping companies actually operate better.
There are a lot of fly-by-night offerings circulating right now when it comes to Ai Services.
For example, 'second brains'.
Throwing scattered data into a second brain while the processes underneath stay broken does nothing. The companies who will absolutely destroy their competition over the next 5 years are rebuilding how they work from the ground up.
Starbucks is showing you what other companies will be doing over the next several years.
Your job is to position yourself to facilitate that process for as many companies as you can.
Lakshman Dhasarathan retweeted
This is a great post if youre thinking about applied AI in the enterprise. The headline of this post is about what companies have huge upside from AI, but the deepest nuggets are about what AI transformation looks like in an organization.
It’s fundamentally about changing the underlying workflow or business process. As we move from chat tools to agents, those agents actually have to be deployed against workflows, which usually span multiple functions in an org. This is a different way of deploying AI than solely rolling it out to end users. It takes much more work upfront, but the results are the things that actually drive significant ROI.
“Software asks the employee to adopt a tool, but infrastructure changes the operating layer underneath the employee. The employee should still know what happened, and the process owner should still be able to pause the workflow, change a rule, approve an exception, or pull a person back in when needed. But the value should not depend on someone remembering to use the AI every day.”
The winners of this will be the platforms that can be deployed for specific workflows and business processes with a deep domain expertise. The playbook will often heavily require FDE support, change management, getting data well organized, be able to have comprehensive evals for the workflows, and much more to get right.
Lakshman Dhasarathan retweeted
𝕏 is a great platform for product announcements, especially if done by the CEO directly. Way more interesting to the public than generic press releases.
This post by Mark Zuckerberg already received over 12 million views for free!
