Shashi ๐ฌ๐ง
@Shashikant86
Followers
2K
Following
1K
Media
849
Statuses
12K
Founder @SuperagenticAI, ex-๏ฃฟ Apple, Building & Optimizing Agentic AI, DSPy, GEPA, SuperOptiX, Organiser London Agentic AI, Researching Quantum AI. Father to 2
London
Joined January 2010
๐ ๐ฆ๐๐ฝ๐ฒ๐ฟ๐ฎ๐ด๐ฒ๐ป๐๐ถ๐ฐ ๐๐ ๐บ๐ฎ๐ฑ๐ฒ ๐๐ฎ๐๐ฒ๐ ๐ฎ๐ ๐ข๐๐ฆ๐ ๐๐ ๐ฆ๐ฎ๐ป ๐๐ฟ๐ฎ๐ป๐ฐ๐ถ๐๐ฐ๐ผ ๐ฎ๐ฌ๐ฎ๐ฑ! ๐ ๐ก @SuperagenticAI made its ๐ฒ๐
๐ต๐ถ๐ฏ๐ถ๐๐ถ๐ผ๐ป ๐ฑ๐ฒ๐ฏ๐๐ showcasing our vision for ๐๐๐น๐น-๐ฆ๐๐ฎ๐ฐ๐ธ ๐๐ด๐ฒ๐ป๐๐ถ๐ฐ ๐ข๐ฝ๐๐ถ๐บ๐ถ๐๐ฎ๐๐ถ๐ผ๐ป and connecting with
1
0
2
First release of DSPy Code focused on basic Python ๐ Virtual Environments venv. to gather versions and context. Next release will be for the UV users who know advanced dependency management with uv so that DSPy Code can generate context and response with speed of Rust and UV. No
DSPy Code comes with even cooler features to connect to local as well as Cloud based models. ๐ฆพDSPy Code can now use power of LLM to generate better respond and v0.1.0 has been released! ๐ Come on, Try DSPy Code Now and build and optimize your DSPy pipelines. #DSPy #GEPA
0
0
2
DSPy Code comes with even cooler features to connect to local as well as Cloud based models. ๐ฆพDSPy Code can now use power of LLM to generate better respond and v0.1.0 has been released! ๐ Come on, Try DSPy Code Now and build and optimize your DSPy pipelines. #DSPy #GEPA
๐ Today, @SuperagenticAI excited to announce: DSPy Code ๐งโ๐ป๐ฉโ๐ป Open Source Library ๐ ๐ฆพ DSPy Code: The comprehensive CLI to build, validate, and GEPA-optimized @DSPyOSS applications. (Claude Code for your DSPy powered agents) โก๏ธ โข๏ธThis is very early stage/ foundation of DSPy
0
0
4
๐กSuperRadar has been updated with, featured tools this week with Opus 4.5 and Gemini3 as highlight of the week. Agent Sandbox from Kubernetes also make some vibes and new Web3 based Agentic IDE InfinityG added to the SuperRadar. Updated SWE and Terminal Benchmarks with latest
1
0
0
If your GitHub README file has following , then I know its AI generated and which models you have used to generate that .. ๐ ------ ๐ Acknowledgments Built with โค๏ธ by the ....blah blah blah โญ Star us on GitHub โข ๐ Read the Docs โข ๐ Report Issues ๐ Documentation ๐ฌ
0
0
0
In this thread, I promised to open source DSPy CLI while watching @aiDotEngineer DSPy/GEPA corner posted @swyx interviewing some guys ( I never heard of them before al least on X) .. "dspy-cli" package was already taken so open sourced DSPy Code. Its very early and might need bit
@swyx @tarunsachdeva @thomastjoshi @aiDotEngineer @DSPyOSS Promise Made and Promise Delivered (One day late) I have open sourced the "DSPy Code" library allows you to build and optimise your DSPy GEPA powered AI Agents! ๐ X Post :
0
1
2
@SuperagenticAI @DSPyOSS ๐ฌ Landing Page : https://t.co/Tn5fOeMcrZ ๐ปGitHub Repo: https://t.co/fLVcg56QAC ๐Docs: https://t.co/JJ59rT7v7o ๐ฃAnnouncement Post:
super-agentic.ai
Full Stack Agentic AI Company
0
1
1
๐ Today, @SuperagenticAI excited to announce: DSPy Code ๐งโ๐ป๐ฉโ๐ป Open Source Library ๐ ๐ฆพ DSPy Code: The comprehensive CLI to build, validate, and GEPA-optimized @DSPyOSS applications. (Claude Code for your DSPy powered agents) โก๏ธ โข๏ธThis is very early stage/ foundation of DSPy
1
2
3
If you are thinking Agent Optimization at highest level then you should definitely watch this talk from Alberto on Meta-ACE from @aiDotEngineer . DSPy, GEPA has setup strong foundation for Agent Optimzation recently we have ACE (Agentic Context Engineering) and now we are talking
2
0
4
Opus 4.5 landed, trie to use it but I no luck, Ran out of credit, just in time but need to ship something that I promised .. Feeling poor to try Opus 4.5 .. The word's best agentic coding model at least for today!
0
0
0
Today, @AnthropicAI launched Opus 4.5 @claudeai and claims the best ever model for Agentic Coding! ๐ปThe race is real for coding models, Let's try Opus 4.5 Now!
Our engineers have found that Opus 4.5 handles ambiguity and reasons about tradeoffs without hand-holding. When pointed at a complex, multi-system bug, it figures out the fix. Overall, Opus 4.5 just "gets it."
0
0
0
Now @antigravity talking about Agent Assisted Development. Not the Agent-Driven Devely yet! Whatโs your view on Spec Driven Development @OfficialLoganK @_mohansolo ?
You can decide how much autonomy you give to the agent in @antigravity Drop asynchronous feedback within artifacts and the agent folds it in when the time is right.
0
0
0
WoW ๐คฉ..I think @dwarkesh_sp is having best podcast guests right now especially in the AI space right now! Great work, Dwarkesh.. look forward your conversation with @ilyasut and donโt forget to ask whatโs ETA for first product of @ssi or at least teaser!
0
0
0
Currently, most of the agents are over engineered from the half-arsed people from Software Engineering background coming to AI/ML applying their crazy SWE ideas to AI/ML. (horrible people) .. Loved the advise from @pashmerepat from @cline at @aiDotEngineer Code Summit NYC is
1
1
1
One of the authentic talk from @aiDotEngineer Code Summit in NYC is for @EnoReyes from @FactoryAI .. No product pitch, no hard sell, just pure knowledge sharing. Product (Factory CLI) speaks for itself. I am so luck to meet Eno, at @_odsc AI in San Francisco.. Great person and
2
0
6
@_odsc ๐ Full Blog Post: https://t.co/a711rEFFEX ๐ฅ YouTube: https://t.co/DdqI76M09t ๐ Slides:
0
0
1
๐ฅVideo of talk (10 min) ODSCI AI San Francisco ๐ Open Data Science Conference (ODSC) AI West ๐ฅ @_odsc ๐กAgent Optimization is the biggest unsolved problem in AIโฆ so why is everyone ignoring it? (Because everyone trying hard to solve but no success yet) They will keep
3
1
1
Visited @AnthropicAI at Cafe Air Mail @claudeai with my little AI champ in London ๐ฌ๐ง. Shout out to the guy second picture, might have been a fan! Missed you @sammcallister by few min but hope we will do it again in London! Nice banana cakes ๐ฐ though!! #LondonAIVibes
I feel like we should do this again This weekend only for the last time? Air Mail Chiltern Street, London First come, first served This weekend only
1
0
37