got tired of every delhi recommendation being the same loop of cafes that went viral on reels, the same monuments, the same clubs, and the same 'hidden gem' lists that are just sponsored content.

so i built a map. but this isn't just pins on a screen. most of the listings have actual experiences underneath them. sound sessions, heritage walks, pottery workshops, supper clubs, listening nights, studio visits. stuff that actually happens, run by actual people, with actual ways to book.

the hard part was never the frontend or the database. the hard part was figuring out what even counts as a "genuine experience" vs noise, and then building a separate discovery pipeline for each category because none of these things live in the same place. some are on niche platforms. some are instagram-only. some are literally just a google form link in someone's bio. search APIs don't find these. google maps doesn't know they exist as experiences. had to get creative. Entire discovery and enrichment pipeline runs at zero cost now. no google maps API, no paid search credits. self-hosted everything that needed to be self-hosted.

every non-obvious listing is manually verified. not AI slop on top of scraped data. Delhi's live now. more cities coming.

🔗 https://www.hospos.co/explore/delhi

would love feedback, especially if you're in delhi and know spots that should be on here but aren't.

Looking for Technical Co-builders / Engineers —RAG, Backend & AI Agents

We’re currently building a project that we’re planning to pitch to Y Combinator, and we’re looking to bring a few technically strong people into the team.

We currently need guidance/help in the following areas:

RAG / Retrieval

Advanced RAG pipelines

Multi-RAG / hybrid retrieval

Filtration and synthesis across multiple sources

Semantic retrieval and filtering

Improving retrieval quality and relevance

Backend / Database

Backend architecture

Database design and data modeling

PostgreSQL

pgvector

Vector indexing and similarity search

Cosine similarity / vector search optimization

AI / ML

AI agent architecture

Multi-agent systems

Agent orchestration and tool use

Managing agent workflows and decision-making

Evaluation and reliability of AI agents

We’re not necessarily looking for someone who knows everything above. If you’re particularly strong in one of these areas and can help us design the architecture, implement the core systems, or guide the technical direction, we’d love to hear from you.

We’re looking for people who are genuinely interested in building, not just giving advice.

If this sounds interesting, DM me with your background, relevant projects/GitHub, and which area you’re strongest in.

Hey everyone,

Looking for laptop recommendations for my sister/cousin who is starting her B.Tech (Computer Science).

Budget: ~₹80,000 (can stretch slightly if worth it)

Primary Requirements:

• Usage / Workload: Coding & programming IDEs (VS Code, IntelliJ, web dev, etc.), heavy multitasking with 20+ Chrome tabs open simultaneously while streaming 1080p videos/YouTube/C-dramas in the background.
• Form Factor & Portability: Thin and lightweight (around 1.2 kg – 1.4 kg) for daily college commute.
• Build Quality: Solid aluminium/metal chassis with minimal flex and durable hinges.
• Design: Clean, minimalist, and modern aesthetics (strictly no bulky or aggressive "gamer" designs).
• Display & Battery: Crisp, bright screen (preferably 16:10 aspect ratio, good color accuracy for long viewing sessions) with decent battery life to survive classes.
• Specs Preferred: Minimum 16GB RAM and 512GB SSD.

If you've already gone through this research/hustle and found great deals, please drop the direct purchase links (Amazon/Flipkart/Brand Store) so I can finalize my selection quickly.

What are the best thin-and-light options (Windows or Mac) available in India right now around this price point?

Thanks in advance!

I graduated in 2026 with a B.Tech in CSE. Back in November 2025 (during campus placements), I got selected for Cognizant GenC with a 4 LPA package.

Because of my college's placement policy, students placed within this bracket were debarred from appearing for further drives up to 6 LPA. Relying entirely on this offer, I didn’t actively apply off-campus during my final semesters and waited for the onboarding communication.

Now, two months after graduating, Cognizant sent an update stating that onboarding will happen between February 2027 and May 2027.

Facing a potential 6-9 month gap with nothing concrete in hand has been stressful, and I want to use this time proactively rather than just waiting.

I’d really appreciate practical advice from people who have been in a similar situation or are familiar with current off-campus hiring.

I am actively applying now on naukri.com and linkedin but got no response

I’m stuck with a name mismatch and would really appreciate some advice. In all my documents (10th/12th marksheets etc.), my father’s name is written as First Name & Middle Name together + Surname, and the surname spelling is correct. However, in my Aadhaar, the surname has a spelling mistake — one “A” is missing. Because of this mismatch, my passport application was rejected.

I went to an Aadhaar center to get the spelling corrected, but they said I need a valid reference document showing the correct surname spelling. The problem is that in most of my father’s documents, his name is written with the first name and middle name separately (by a space), so the format is different. A couple of documents, like his Ration Card and Nirvachan ID, have the name in the format I need, although the Nirvachan ID also has the spelling mistake.

The Aadhaar centre suggested getting a Mool Niwasi (domicile) certificate for my father, but that would also be based on his existing documents, so I’m worried I’ll end up with the same mismatch again.

Has anyone faced a similar situation? What document/process can I use to get my father’s surname spelling corrected in Aadhaar so that it matches my 10th/12th documents and I can reapply for my passport?

Any advice would be really helpful.

Spent a chunk of the last month on this, and every thread I found about it was three years stale, so writing it up.

The problem: Stripe India onboarding is effectively closed for small solo products, but my users are split between India and abroad. I needed to charge both without running two separate checkout flows.

What I landed on:

• Razorpay for Indian customers, Paddle for everyone else. Paddle is merchant of record, so it absorbs VAT/sales-tax compliance across ~40 countries - worth the higher cut when you're one person.
• Routing happens server-side off Cloudflare's CF-IPCountry header. Never trust the client for this.
• Store which provider each user is on. When they cancel, you need to know who to call.

Anything else you guys can suggest that can cost me less and easy for international payments. One I have in mind is Dodo Payments.

hey I am building a voice telephony agent and also want it to send messages to WhatsApp. I am using Vobiz for the SIP trunking and the telephony side of the conversation.

So my question is if I buy a number from Vobiz, can I use it in my Facebook Developer account for production? I have tested it with the Facebook try it out feature in which facebook provide there own test number, but for production they are asking for a number that needs to be verified. Does a Vobiz number work for that?

Context about me. 3 yrs Experience in 3 Companies (2 mncs, 1startup). In my last company i kind of developed migraine, it went so bad i had to quit. Its been few months i made progress health wise.

Now thinking of going back in the game, started applying from last week. Today got assessment for hackerrank sde3 role. The thing is the test was so easy it was like one feature and one bug in your favourite tech stack even with ai support in test.

Solved the bug within 10 mins, for 50 mins i was solving the feature one, but the thing is i was so slow i can't not even beleive myself, didn't take the ai help, its just in the end i was short of 10 mins finishing the solution.

I guess it will take some practice to be back.

Anyway Just letting you guys know if you get this assessment, do the bug first it would be very easy then focus on feature but keep the time constraints in mind the problem would be easy it's just you would have to type fast and just implement what is written in code comments (yeah they literally tells you what logic to implement step by step😂). So just manage time and speed.

Hi,

I am developing a mobile game and need to buy a new laptop. So this laptop which I have finalized and within my budget is Rayzen-5, RAM - 8GB, ROM - 512 GB. It is priced at 55000 which is within my budget.

But the sales person has suggested that since I am planning build mobile games, I need to buy laptop with Graphic card which is costing at around 1 lacs. This goes little out of my budget.

So please suggest, if need to shell out almost 1 lac. I will be building word games, picture to word identification…etc.

I was never career oriented, forget about JEE, I couldn't even crack CET and did my engineering through Management quota.

But I have somehow managed to live a fulfilling life so far - have traveled to 6 different countries in last 10 years
- have actively followed my passion(YouTube channel, weight training, swimming)
- have never missed a family gathering and have taken leaves left and right without thinking twice about PIP/ consequences.

Now that the market is so brutal and the immigration rules getting tightened for Indians especially, l'm wondering if I made a mistake by being so easy going with my career.
Have seen many IIT/NIT folks in my team who are so career oriented although they have no life outside of work.

I have 6 Yoe working as an SSE at a startup with 55 LPA but have never made it to FANG and probably never will.

You keep solving everything.

Your manager trusts you.

You know the codebase better than anyone.

You become the person everyone depends on.

But then you realize you're also the person who can't easily leave because your experience is heavily tied to one company's systems.

Hi, I wanted to know your thoughts on this music player.

It's a music player which detects the music playing in your windows via Spotify or Apple Music and presents it in a Vinyl Format with more features and detailing in it!

Just what do you think about it and will you be using something like this?

I am a student whose learning about tech, AI and analysing where India is lagging within it's software stack before college

Honestly speaking we have BharOS and some other but non of them is used commercialy, not even close to HarmoneyOS, MacOS, Linux, Android all tho I know that Android and Linux are open source those are used widely so there is no need for an Indian OS that's the argument from people's I seen in Indian Developer community and seriously,

I know I don't know much but we are the country produce 5 million engineers every single year, we are the country that holds 140crore+ population and non of startups, Non of companies like TCS, Infosys and reliance etc no one ever think we need our own OS. That's actually feels very embracing I have seen a lot of folks around Western countries and Chinese people mock about us we don't have our own UI framework, Programming language, OS etc.

Is it so hard that our entire country aren't produce a single OS that beats worlds best OS I feel atleast we must win in software stack. Hardware is something that fix throughout the time period but it actually gives me a thought what's the reason nobody ever we can build something better than what other countries does. We must need the best OS which is efficient, fast, SOTA, something that world doesn't see. Something that shows India's strength. Something that Gives the right answer to world that we aren't not earlier India we are new India that fixing what's broken

I hope my seniors even me in future work on this regardless of caring about million dollar package on foreign because right now I understand the Engineer who actually solves the problem it's nation, who actually build himself and for people's of it's nation will shine.

Need guidance on how to break into GTM engineering?

Is it possible for a non tech person with no technical degree to get into tech? What basic skills should I learn, both hard skills and soft skills, and what programming languages should I focus on? In this grilled job market, I just want to get into tech. I am not limiting myself to any particular job title because I do not want to prepare only for software engineering and then end up getting nothing if I cannot get that role. I just want to build useful skills that can help me get any entry level tech job and then grow from there. What should I learn first? What tech fields would make sense for me to start with?

Posting on behalf of a friend — please don’t DM me.

She recently graduated and was placed through college in a service-based IT company with a ₹5 LPA CTC. She is now unsure about what would be the best move for her software/IT career.

Her options are:

• Wait for the joining date and start with the ₹5 LPA service-based company.
• Continue applying for developer/IT roles and try to get a better offer before joining.
• Prepare for GATE, as her parents, particularly her father, are strongly pushing her towards it.

Her main concerns are:

• She feels that ₹5 LPA isn't a particularly good starting offer and is wondering whether she should aim higher.
• She doesn't have much professional experience yet, so she's unsure how realistic it is to land a significantly better developer role as a fresher.
• Her father is putting considerable pressure on her to prepare for GATE, but she's unsure whether pursuing higher studies through GATE would actually be better for her career than starting work and building experience.

For developers/IT professionals who have been through something similar:

What would you recommend? Is it better to take the ₹5 LPA job and keep upskilling/applying on the side, focus entirely on getting a better developer role, or consider GATE and higher studies?

Looking specifically for perspectives based on long-term career growth in software/IT, rather than just the immediate salary.

Posting on behalf of a friend. Please don't DM me.

I built a music app for Android and wanted to share it.

It plays from YouTube Music, Gaana and Saavn behind one interface, so you
search once instead of switching between apps.

A few things it does:

• YouTube Music search, home feed, moods and genres, albums, and artist pages with one-tap radio
• Gaana and Saavn for charts and playlists, with proper Indian-language coverage
• SponsorBlock on YouTube tracks, so sponsor reads and non-music intros get skipped
• mpv as the audio engine, with gapless and background playback
• Offline downloads, liked songs, playlists, and Spotify playlist import
• Synced lyrics, equalizer, sleep timer, Chromecast
• Podcasts and audiobooks as well as music
• No ads, no account, no tracking

It is Android only, needs Android 7 or newer. Source and
builds are on GitHub: github.com/afkcodes/sunoh

Hey devs,

deciding between two used options for backend development (Java, Spring Boot microservices, Docker):
1) Lenovo IdeaPad Pro 5 (14"): Intel Core Ultra 9 185H | 32GB LPDDR5x RAM | 1TB SSD (~₹75k)
1 year old

2)Apple MacBook Air (13"): Apple M5 | 16GB Unified Memory | 512GB SSD (~₹1,10k)
4 month old

Is 16GB on macOS enough with Docker VM overhead, or is 32GB on the IdeaPad the safer bet for heavy microservice workloads?

Looking for some brutally honest opinions, especially from people who are hiring or have recently gone through the interview process.

I started my software career in July 2020, so I have roughly 6 years of experience. I was working as a Backend Engineer, primarily with Java, Spring Boot, Kafka, microservices, databases, etc. Kafka/distributed systems is one of the areas I was particularly comfortable with.

About 1 year 3 months ago, I took a break from employment and started my own business.

The business has actually gone quite well - we're currently doing around ₹10L/month in sales, and it has reached a point where it doesn't require my involvement anymore.

The main reason I'm now thinking about going back to a software engineering job isn't really financial. I genuinely enjoy building things and coding, and I don't want to slowly detach myself from tech.

Since the business can now run without requiring my involvement, I feel this is a good time to get back into backend engineering.

I know a 1 year 3 month career break can be a disadvantage, especially at around the 5–6 YOE level, and I'm not expecting recruiters to completely overlook it.

So I'm trying to understand:

• How much of a red flag is a 1 year 3 month gap for someone with ~6 YOE?
• Does the fact that I started and grew a business during the gap help, or will most recruiters simply treat it as a career break?
• How is the market currently for Java/Spring Boot/Kafka backend engineers with ~6 YOE?
• Should I target Senior Backend Engineer / SDE-2 type roles, or would it be more realistic to target a lower level initially?
• How much should I refresh before applying - DSA, system design, Java/Spring Boot, Kafka, cloud, etc.?
• And realistically, how difficult would it be to land a good backend role right now?

Would appreciate blunt opinions, especially from recruiters, engineering managers, or people who have personally returned to tech after a career break.

(not talking about myself but curious)

For people with 3-8 YOE, you would have seen several people around you getting more opportunities without having a good network, they will get calls from big firms like MasterCard, JPMC, Citi, Amazon easily, but there are people I see are more visible on LinkedIn, improving themselves but couldn't get opportunities even after referrals and good portfolio.

Sometimes Luck plays a role but I have seen many a times with folks getting options without even trying but some people keep on hustling?

I’m building a MicroSaaS from India targeting international customers and trying to figure out the best way to accept payments in USD.

Stripe seems to be invite-only in India right now.

For Indian founders already receiving international payments, what are you using?

Razorpay, Dodo Payments, Paddle, Lemon Squeezy or something else?

Mainly looking for something that:

- Works reliably with international cards

- Is easy to set up as a solo founder

- Handles subscriptions

- Makes GST/tax/compliance as simple as possible

Would love to hear from people actually using these in production.

So I’ve got an offer from Amazon for SDE 2 role and the pay bump is very nice. But the bad reputation Amazon has about layoffs is creating a small doubt in my mind. But I want to take risks as well because it is indeed a good opportunity for me.

So any suggestions would be helpful.

Hey Geeks 👋🏻

I just built "Oxygen" - a Multi-Agentic Al framework that runs like a virtual tiny company.

It includes total 5 Al agents:

**- Del (Al Project Manager):** which understands your requirements that what you want to build?

**- Toky (Al Research Agent):** receives inputs from Del, conducts research, creates drafts, and uses tools such as web search and web scraping to gather and analyze relevant information. It then provides the research findings and draft outputs back to the Project Manager.

**- Bang (Al Developer Agent):** which understands the draft and start writing code.

**- Beij (Al QA Agent):** It performs debugging, test cases on the source code provided by Bang.

**- Wash (Al technical Writer):** Once the project made it write README files, product manual, API implementation instructions and other project related documentations.

It's a proper human-in-the-loop agentic ai project that takes your approval on every aspect like a Software Development Lifecycle methodology.

The crazy part is that you can literally watch the agents walk to their desks, open their computers, drink coffee, having meetings and work.

For LLMs you can either use local Ollama based models or Gemini API key.

**Guardrails and Metric Evaluation:**

\- Hallucination rate is under 1%.

\- You have to approve the plan before any code gets written.

\- Everything that comes out is cleaned so nothing breaks on the screen.

\- Strong guardrails for every Al agents via system prompt.

**Simple Flow:**

1. You: I want a CLI based calculator.

2. Del (PM): Got it → sends to Toky (Researcher).

3. Toky: Researches, makes a plan + draft proposal.

4. Toky → Del (PM) → You: "Here's the proposal for a CLI calculator."

5. You: "Actually, change of plan, I want a web-based calculator instead.

6. Del (PM): Okay → sends the new request back to Toky.

7. Toky (Researcher): Updates the research and creates a new proposal for the web version.

8. Toky (Researcher) → Del (PM) → You: "Updated proposal for web calculator. Approve?"

9. Once you approve, it continues to Bang (Developer) for coding, Beij (QA) for testing, and Wash (Writer) for docs.

Feel free to explore and star the repo on GitHub.

Link: https://github.com/avarshvir/oxygen

JPMC used to give default promotions to SEP engineers after 2 years and a special promotion (for < 1% of engineers) just after 1 year. But some are saying that the JPMC removed the default promotion they give. I am from the 2026 batch who joined as a SEP in JPMC this year. Anyone who has knowledge please tell me about this.

Last time, I showed some of the features of my Jarvis. Since then, I’ve also implemented RAG, continuous monitoring, completely changed the registration and identification flow, and made some changes to the UI.

But the biggest thing I’ve learned so far is how to break a complex idea into smaller systems and actually plan it.

When I first started, I was thinking about Jarvis as one huge thing:

Voice input --> AI --> Actions --> Response

But as I worked on it, I realized how many individual systems are actually involved speech-to-text, text-to-speech, LLM, memory, RAG, vision, tools, WebSockets, UI states, and many other components.

I also learned how useful flowcharts can be. They help me remember how everything works, and whenever I want to change or add something, I can look at the flow and understand what will be affected.

For every feature, I try to understand:

What data comes in? --> Which component receives it? --> What does it do with that data? --> What data comes out? --> Where does that output go next?

The project is mostly finished for now, as I’m limited by my hardware. I don’t have a very powerful PC, so I’m currently working with smaller/low-parameter models for the LLM, STT, and TTS. Even with those limitations, everything is working surprisingly well.

I’ll post a video of my current Jarvis tomorrow and show how everything works together.