Full stack developer | JavaScript | Node.js | Express.js | Sequelize | React.js | Python | Typescript | SQL | NoSQL | AWS Ec2 | AWS S3

Joined September 2024
23 Photos and videos
Deepthi Purijala retweeted
Doubt yourself every day. Imposter Syndrome is real and essential. We keep looking for resources to overcome it, but to be honest, there is no real need. I let it hit hard, as it, pushes me to learn more, dig deeper, and explore concepts I would have otherwise overlooked. But here's the critical part - this self-doubt should not erode your confidence as too much of it can devastate you completely, so you need to maintain a balance. So, every time you solve a problem, fix a bug, or implement a feature, take a moment to acknowledge it and reward yourself. Remember, even most senior engineers face challenges that they are not equipped to handle. But, the difference is, they use their self-doubt as a push toward new learning opportunities. Did I ever feel like an impostor in my career? yes; do I feel it today? absolutely. There were moments when I felt like an imposter and thought it was way over my head. But that feeling pushed me to learn more, work harder, and dig deeper. Remember, the goal is not to eliminate self-doubt but to leverage it to find the growth areas and work on them. Hope this helps.
32
62
936
28,379
Crossed 75 DSA problems on LeetCode. Still a long way to go, but consistency is starting to compound. On to 100 🚀
2
22
Diagnosed with ADHD. Written off early in life. Michael Phelps turned every doubt into discipline and became the greatest Olympian ever !! “I hate to lose more than I enjoy winning.” — Michael Phelps
1
20
If even Andrej Karpathy expresses such a sense of urgency and adaptation pressure, it serves as a clear signal for the rest of us. The pace of change demands serious, sustained effort. How seriously should we take this moment and how intensely should we grind to stay relevant?
I've never felt this much behind as a programmer. The profession is being dramatically refactored as the bits contributed by the programmer are increasingly sparse and between. I have a sense that I could be 10X more powerful if I just properly string together what has become available over the last ~year and a failure to claim the boost feels decidedly like skill issue. There's a new programmable layer of abstraction to master (in addition to the usual layers below) involving agents, subagents, their prompts, contexts, memory, modes, permissions, tools, plugins, skills, hooks, MCP, LSP, slash commands, workflows, IDE integrations, and a need to build an all-encompassing mental model for strengths and pitfalls of fundamentally stochastic, fallible, unintelligible and changing entities suddenly intermingled with what used to be good old fashioned engineering. Clearly some powerful alien tool was handed around except it comes with no manual and everyone has to figure out how to hold it and operate it, while the resulting magnitude 9 earthquake is rocking the profession. Roll up your sleeves to not fall behind.
1
20
Just watched @akshaymarch7 's "If I Started to Code in 2026" – brutal truth bomb. Escape the middle-class trap or get replaced! If you're serious about staying relevant in 2026 , this is a must-watch. Link: youtube.com/watch?v=lAf8RZh5… What’s your biggest takeaway? 👇
1
8
1,402
1/13 RAG Explained Simply – A Thread 🧵 Most people think: “To make a chatbot know my data, I have to retrain the model every day.” That’s expensive, slow (24h delay), and read-only. RAG (Retrieval-Augmented Generation) fixes all of that! Let’s break it down 👇
1
1
51
10/13 Popular tools • LangChain / LlamaIndex → everything in ~5 lines • But heavy abstraction You can also build it yourself in <100 lines with: • sentence-transformers (embeddings) • FAISS / Chroma / Pinecone (vector store) • Any LLM (OpenAI, Claude, local)
1
1
24
11/13 Simple RAG chain Loader → Text Splitter → Embedding Model → Vector Store → Retriever → LLM That’s the entire architecture.
1
1
28
1/13 RAG Explained Simply – A Thread 🧵 Most people think: “To make a chatbot know my data, I have to retrain the model every day.” That’s expensive, slow (24h delay), and read-only. RAG (Retrieval-Augmented Generation) fixes all of that. Let’s break it down 👇
1
1
31
12/13 Note: Not every RAG needs embeddings! Tiny docs → keyword search or simple splitting works. But embeddings = true semantic search → way better results.
1
6
13/13 Why RAG is winning • No retraining • Always up-to-date • Cheaper • Keeps your data private • Full control over retrieval Every serious app will have a RAG layer soon. #RAG #LLM #GenAI #AICoding
8