Cimpress India

Day 34 of building in public 🚀 Today I focused on adding more polish and motion to the landing page of my Job Tracker app. Most of the core app functionality is already working: dashboard, Kanban, analytics, settings, dark/light mode, localStorage, modals, and toast notifications. Now I’m improving the landing page so it feels more like a real SaaS-style product, not just a static portfolio screen. 🔧 What I worked on: • Added subtle Framer Motion animations to the Hero section • Animated the Dashboard Preview card when it enters the screen • Added stagger animations to the Features section cards • Added stagger animations to the Built With section cards • Animated the CTA section with smooth fade-in • Added hover and tap animation to the CTA button • Improved dark mode consistency across landing page sections • Added small hover interactions to preview cards for a more polished feel ⚠️ Challenges I faced: • Making animations feel smooth without overdoing them • Replacing normal div elements with motion.div carefully • Matching opening and closing tags correctly • Keeping animations reusable with variants • Making sure dark mode, responsive layout, and animations work together 💡 Learning of the day: Good frontend polish is not about adding random animations. It is about using motion to guide attention, improve flow, and make the UI feel more natural. Today I learned how Framer Motion can make a landing page feel more premium when used carefully and consistently. The project is slowly moving from “functional app” to “portfolio-ready product” ✨ (Attached demo 🎥) #LearningInPublic #100DaysOfCode #FrontendDeveloper #ReactJS #BuildInPublic #TailwindCSS #FramerMotion #UIUX

Thrilled to welcome Rohan shetty to the Cargill family! It’s always exciting to see talented individuals join the team and bring fresh perspectives, energy, and ideas. Wishing you a great start and an amazing journey ahead with us. Looking forward to seeing the impact you’ll make! #WelcomeToTheTeam #NewJoiner #TeamCargill #NewBeginnings #Hiring #TalentAcquisition #PeopleFirst #CargillDTDCareers #GrowingTogether

Day 32 & 33 of building in public 🚀 For the last 2 days, I focused on polishing the landing page and making it feel more like a real production-level frontend project. After building the main sections, I started working on the small details that improve the overall user experience. 🔧 What I worked on: • Improved landing page navigation for desktop and mobile • Fixed navbar alignment using a responsive layout • Added mobile menu with open/close state • Created a compact dropdown menu for small screens • Added highlighted background when the mobile menu is open • Connected About, Features, and Built With links to the correct sections • Added smooth scrolling for landing page navigation • Added scroll spacing so sections don’t hide behind the sticky navbar • Added dark mode support step by step across landing page sections • Updated Navbar, Hero, Dashboard Preview, Features, Built With, and CTA sections • Improved CTA button styling and fixed small spacing issues ⚠️ Challenges I faced: • Keeping desktop navigation perfectly centered • Avoiding duplicate dashboard buttons on smaller screens • Making the mobile dropdown look compact and clean • Handling dark mode colors without breaking the light mode design • Fixing small Tailwind utility issues that affected alignment and spacing 💡 Learning of the day: Responsive design is not just about making things fit on smaller screens. It is about making the experience feel intentional on every screen size. These 2 days taught me that small UI details like dropdown width, sticky navbar spacing, hover states, smooth scrolling, and dark mode colors can make a portfolio project feel much closer to a real product. The landing page is now becoming cleaner, responsive, and more polished ✨ (Attached demo 🎥) #LearningInPublic #100DaysOfCode #FrontendDeveloper #ReactJS #BuildInPublic #TailwindCSS #UIUX

Happy to share that I’ve earned my AWS Certified Developer – Associate certification! ☁️🚀 Excited to keep building and learning in the cloud. Verify here: https://lnkd.in/ghkw8PGE #AWS #CloudComputing #Certification

There's a gap between business teams and data teams that nobody talks about enough. Business stakeholder needs a quick answer: "How many customers from Jaipur ordered more than 3 times in the last 90 days? I have a regional pitch in 2 hours." They raise a request. Data team picks it up. But there's always a TAT — other priorities, queue, context-switching. By the time the answer comes back, the meeting is over. The need has expired. Effort wasted on both sides. And even when the data does come back in time, there are follow-ups: "Can you break this down by product category?" "What about the last 180 days instead?" "How does this compare to Lucknow?" Each follow-up goes back into the queue. The data analyst isn't sitting idle waiting — they've moved on to something else. The back-and-forth stretches a 5-minute question into a 2-day thread. The solution has to meet business teams where they already are — not another dashboard they won't check, not a SQL editor they can't use. It has to be something they're already comfortable with. For most teams, that's Slack. So I built SlackQL — a Slack bot where anyone can type: /ask How many customers from Jaipur placed more than 3 orders in the last 90 days? ...and get actual database results back in seconds. Follow-up question? Just ask again. No queue, no TAT, no context lost. How it works: → Slack command hits a Python backend → AI Agent translates the question to SQL using live database schema as context. → Query executes against PostgreSQL with safety guardrails → Results are formatted and posted back to Slack The safety layer was critical to get right — you can't just let an LLM run arbitrary SQL against production data: - Read-only access only — no writes, no deletes, ever - DML keyword blocking as a second defense layer - PII data masked - 100-row result cap and 30-second query timeout - Dynamic schema injection so the LLM only generates queries against tables that actually exist. This doesn't replace data analysts. It handles the interrupt-driven ad-hoc requests so analysts can focus on the work that actually needs human thinking — modeling, deep analysis, data quality. The whole thing is open source: https://lnkd.in/gpskTGDn Tech stack: Python · Slack Bolt · Gemini API · PostgreSQL · FastAPI #DataEngineering #AI #LLM #Python #OpenSource

Reposting for my network.