Prathamesh Sirdesai
@SirdesaiEXE
Followers
108
Following
474
Media
161
Statuses
830
Building @MaximalStudio π , consumer apps, AI tools
India
Joined April 2020
22 & figuring life out by building cool things. π¨βπ» Building MVPs and cool stuff π€ Making AI tools that actually help people π Dream: Build stuff that changes lives π Currently: Learning, building, sharing. One day at a time
0
0
2
Yesterday my @rapidobikeapp rider was listening to a Node.js tutorial. Upon asking, he told me heβs already a developer and is preparing for a switch. Crazy grind & a @peakbengaluru moment for me.
31
92
3K
Cursor indexes your workspace (AST, imports, symbols, dependency graph) and feeds that into the LLM so it can reason about your project like a real dev β not just a text model. π https://t.co/u7h2fM3U5m
#CursorAI #LLM #AIEngineering #ContextRetrieval #DevTools
cursor.com
Documentation for Context
0
0
0
Built a Go-based Sync Service that keeps Postgres βοΈ Elasticsearch always in sync β‘ β
Auto detects DB changes π Has a DLQ for failed syncs π Dashboard + metrics included #Golang #Backend #Elasticsearch #PostgreSQL #SystemDesign
0
0
6
Built a Go-based Sync Service that keeps Postgres βοΈ Elasticsearch always in sync β‘ β
Auto detects DB changes π Has a DLQ for failed syncs π Dashboard + metrics included #Golang #Backend #Elasticsearch #PostgreSQL #SystemDesign
0
0
5
π
0
0
1
One thing I'm grateful for is having 2β3 good friends who guide me whenever I'm stuck.
0
0
0
sometimes life feels motionless, but when the challenges finally arrive, they all come in a flood.
0
0
0
maybe the heavy school bags were the lightest burdens we ever carried..
28
10
260
Software engineer career path: - Intern ($50k) - Junior ($80k) - Senior ($150k) - Staff ($250k) - Farmer ($30k but actually happy)
201
541
12K
today, a classmate saw a folder on my laptop named "models" and thought it was filled with pictures he was disappointed looking at .pkl files...
0
0
0
Integrated Web-LLM into my Mentor-AI app. The model is streamed οΏ½οΏ½οΏ½ cached in IndexedDB β executed via WebGPU in-browser. It converts the client into a local inference runtime with zero server dependency post-load. Tradeoffs: VRAM usage, load latency, and token throughput.
0
0
1
making your own food (diet) daily is next level satisfaction!
0
0
0