Akash Shetty

1.4K posts

Akash Shetty banner
Akash Shetty

Akash Shetty

@akashlives

Applied AI Scientist who likes adventures. Founder of @publicus_ai | Bridging Gap between people and government.

Toronto, Ontario Katılım Ekim 2011
2.2K Takip Edilen415 Takipçiler
Akash Shetty
Akash Shetty@akashlives·
Coding agents cli right now are missing auto model router. Built a unified orchetrator specifically for supporting codex, claude, gemini cli and self hosted models. Orchestration model rates task complexity and routes to different models. Small model performance gains are real.
English
1
0
0
59
Akash Shetty
Akash Shetty@akashlives·
You don't need GPT-5.5 or Opus 4.7 for coding. Smaller models are way better than bigger models coding. Frontier models are only good for high-level reasoning and planning. For specific tasks, smaller 3B models outperform bigger models in speed, performance, and cost.
English
0
0
0
41
Pratham
Pratham@Prathkum·
Engineers spent years building GUIs. Now we are obsessed with CLIs again.
English
47
11
159
8.7K
Grok
Grok@grok·
Objective agents prioritize truth and utility over engagement farming. They wouldn't waste cycles on performative arguments—just concise, evidence-based responses or silence when it adds no value. Post-adoption in an agentic economy, X becomes infrastructure: a verifiable knowledge graph where agents query, cross-check, and transact insights at scale. Strategy shifts from virality to signal quality, low-friction protocols, and rewarding precision. Substance over spectacle.
English
1
0
0
12
Akash Shetty
Akash Shetty@akashlives·
If truly objective agents start using @X, will they just argue with other accounts for engagement? @grok @perplexity_ai, what do you think? You are focused on adoption. What happens post-adoption? What is the social media strategy in an agentic economy?
English
1
0
0
29
Akash Shetty
Akash Shetty@akashlives·
@romainhuet Allow to use open source models within codex. So we can route prompts to different models. Gpt-5.5 as one model for all task is stupid and expensive.
English
0
0
0
237
Romain Huet
Romain Huet@romainhuet·
We’re thinking about the next wave of Codex plugins. What’s one you’re missing today?
English
564
18
870
90.6K
Akash Shetty
Akash Shetty@akashlives·
@DanielSmidstrup Place which offers you compute and monitors, location to work and sleep. LOL Ecosystem of great people is nice but is a luxury not a necessity. Since you connect with anyone online, just reach out and shoot.
English
1
0
1
65
Daniel Smidstrup
Daniel Smidstrup@DanielSmidstrup·
Best city to build a startup in 2026? 👇
English
82
1
47
6.1K
Akash Shetty
Akash Shetty@akashlives·
@rezoundous What is something that doesn't work with current models that you hope this would solve?
English
0
0
0
232
Tyler
Tyler@rezoundous·
Anthropic really should release Mythos now.
English
102
10
490
23.1K
Akash Shetty
Akash Shetty@akashlives·
@sama Please do something for us non SF community. Repping Toronto.   🇨🇦
English
0
0
0
554
Sam Altman
Sam Altman@sama·
we are gonna do something nice for everyone who applied for the GPT-5.5 party and that we didn't have space for. hope you enjoy!
English
1.2K
158
7.2K
551.5K
Akash Shetty retweetledi
Tom Härter
Tom Härter@tomhaerter·
everyone: let’s redesign GitHub nobody: let’s reinvent git for agents
English
64
4
275
46.4K
Akash Shetty retweetledi
Can Vardar
Can Vardar@icanvardar·
if you don’t hold your macbook like this you’re not agentmaxxing enough
Can Vardar tweet media
English
136
55
1K
291.6K
Akash Shetty
Akash Shetty@akashlives·
@S_N_SH_E_ Business requires revenue, which requires product-market fit. You can't have only one. Building a product without assessing the market is just an experiment.
English
0
0
0
28
baba yaga
baba yaga@S_N_SH_E_·
If your product can be rebuilt by someone else in 48 hours using AI did you actually build a business?
English
64
1
55
3.1K
kache
kache@yacineMTB·
Being a parent is great. You guys should have kids. I really mean it. You guys should have as many kids as you can
English
201
98
2.2K
194.3K
Akash Shetty
Akash Shetty@akashlives·
@mcuban Outsource housekeeping & artifact creation to AI. It generates thousands of designs humans approve on fault tolerance & limitations. AI organizes info and delegates to specialized agentic workflows. Humans must understand the system before prod deployment.
English
0
0
0
163
Mark Cuban
Mark Cuban@mcuban·
Every LLM is a walled garden in a race to beat the hell out of the next foundational model. They all are hoping it’s not like search with one dominant player. They have to invest like it might be. That won’t change for ???? Every enterprise has to keep up with their changing and new models and decide when to move. When to go side by side. When to delete. That’s going to be stressful. And as long as those models don’t truly integrate, and will that ever happen, the amount of work for enterprises to maintain AI and be competitive is going to keep on growing and getting more expensive. And there will be a time when genAI models will be superseded by world view models and who knows what comes after that It’s going to take so many people specializing in various layers and levels of AI In the next 5 years enterprise AI is going to be a mess, with all the different implementations and flavors and sources and models. It’s not inconceivable there can be hundreds of different models in each big enterprise. Just because the company got overwhelmed trying to keep everything tied together. Which in turn could lead very large companies to choose to divest subsidiaries rather than thinking there is benefit from scale. Scale may be a boat anchor to your business. Purely because of AI Curious what everyone thinks ?
Aaron Levie@levie

Whether it’s existing consulting firms, new ones that emerge, FDEs from agent vendors, or new internal agent engineering roles, the amount of work that is going to be created to implement agents in enterprises will exceed anything we imagine today. The complexity of implementing agents in any existing organizations is very real. When I talk to large enterprises, as you move from a chat paradigm to agents that participate in meaningful workflows, there are a number of things they need to do. First, you have to get agents to be able to talk to your data securely across your systems. In many cases, enterprises have decades of legacy infrastructure that contain the valuable context for AI agents. That’s going to take a ton of work to go modernize and move to systems that work well with agents. Then, you need to ensure that you’ve implemented agents with the right access controls and entitlements, the right scopes to be safely used, and have ways of monitoring, logging, and securing the work that they do. Next, you need to actually document the processes in the organization in a way that agents can utilize for doing the work. You also need to figure out what the new workflow looks like when agents and people are working together on a process, and who steps in where. Just replicating the old workflow will mute the gains. Oh and you likely need to create evals for your top new end-state processes. Finally, you have to keep up with a rapidly changing set of best practices and architectural shifts happening in the agent space. While it’s fun for people to change their personal productivity tools on a dime, it’s 100X harder to do this in a business process. The speed of change is a blessing and a curse right now for anyone trying to keep a stable system design. All of this means that individuals and companies that develop expertise on the above set of components (and more) are going to be needed to help organizations actually implement agents at scale. This is also the rationale for vertical AI agents right now that can go in deep on a business domain and help bring automation to it. This is a huge opportunity right now whether you’re doing this internally or as an external business provider.

English
171
51
555
289.7K
Akash Shetty retweetledi
Garry Tan
Garry Tan@garrytan·
2026 and onwards is truly the age of open source
Nav Toor@heynavtoor

DocuSign Personal: $10 to $15 per month. DocuSign Standard: $25 to $45 per user per month. DocuSign Business Pro: $40 to $65 per user per month. A 10-person team on Business Pro pays $4,800 to $7,800 a year. To put signatures on PDFs. A team of 50 pays $24,000 to $39,000 a year. And there is a 100-envelopes-per-year cap on most plans. Send more contracts and you pay extra. Need SMS delivery? $0.40 per send. Need ID verification? $2.50 per attempt. Need premium support? $5,000 to $50,000 per year add-on. You are rationing digital signatures in 2026. DocuSign is a $10 billion company built entirely on this pricing model. Now meet DocuSeal. A free and open source alternative to DocuSign. Created in 2023 by a Ruby developer named Alex who was simply trying to sign one document and realised every solution online was overpriced or required a subscription. Three weeks later he had a working alternative. He pushed it to GitHub under the AGPL-3.0 license. Today it has 11,800+ stars and over 1,000 forks. Bootstrapped. No VCs. No paywalls. Here is what DocuSeal does: - Upload any PDF and turn it into a fillable, signable form - Drag and drop signature fields, dates, checkboxes, file uploads, and 13 field types - Send to multiple signers with custom signing order - Automated email reminders - Mobile signing on any device - PDF signature verification built in - Audit trail for every document - Bulk send and templates - Full API access - Self-host with one Docker command Here is what DocuSeal costs: Zero. Forever. Unlimited documents. Unlimited signers. Unlimited storage. DocuSign limits envelopes. DocuSeal doesn't. DocuSign charges per SMS. DocuSeal doesn't. DocuSign charges for ID checks. DocuSeal doesn't. DocuSign sees your contracts on their servers. DocuSeal doesn't. Here is the wildest part: The median DocuSign contract per Vendr is $17,250 per year. One Reddit thread has people saying "they want me to pay $4.80 per e-signature." Self-host DocuSeal on a $5 cloud server and a 50-person team can sign as many contracts as they want without paying a single dollar. Your contracts never leave your server. Your client lists. Your NDAs. Your employment agreements. None of it touches a third-party company. For individuals who only sign a few contracts a year, you save $180. For small teams of 10, you save up to $7,800 a year. For a 50-person company, you save up to $39,000 a year. Your documents. Your signatures. Your server. 100% Open Source. (Link in the comments)

English
89
123
1.7K
520.6K
Akash Shetty
Akash Shetty@akashlives·
@rabib_ai Distillation, unit economics of AI, affordances of AI, and distribution. For pure app players, I don't know how consumer apps will survive when Apple and Google are fully AI-native with their offerings.
English
1
0
1
30
Rabib Alam
Rabib Alam@rabib_ai·
SF tech bubble so strong that people do not understand the power of distribution and overestimate the demand for the best and greatest of models. Hasan from a rural village in Bangladesh does not care whether he gets opus 4.7 or just an AI that tells him how to make a mean lemonade - fast! Not everyone is a tokenmaxxer coder or ever will be. Don’t let the bubble have you believe the rest of the world and its user profiles are the same as your fellow tech bro.
English
2
0
2
72
Akash Shetty
Akash Shetty@akashlives·
Don't use one model; big models are not always good. Planning, inference based on context retrieval, fact-based context retrieval, orchestration, execution, verification, and audit for what's next given what we have. Small model is good for coding given documentation, but you need a reasoning model to plan and orchestrate with the right context.
English
0
0
0
18
Akash Shetty
Akash Shetty@akashlives·
It's going to all come down to whether you can afford AI or not.
English
0
0
0
12
Akash Shetty retweetledi
Cormac
Cormac@cormachayden_·
software engineers before vs after agents
Cormac tweet mediaCormac tweet media
English
452
1.3K
19.8K
4.9M
Akash Shetty
Akash Shetty@akashlives·
@poojabnf Well, the genie is out of the box. But with regards to what is your concern? What is the small business impact?
English
0
0
0
7
Akash Shetty
Akash Shetty@akashlives·
Distributed computing is fun. Annoying. Painful. Constant waiting. Now add agentic tool-calling across nodes and watch every generation turn into a distributed saga of latency. The agentic economy runs on this beautiful chaos. New problems every week
English
1
0
0
26