Devectra / Work

Evidence over mockups.

Real systems built for real organizations. Each one solved an operational problem, not just a design brief.

01 / 05

Civil Engineering

SETU — Civil Tendering Platform

Live site ↗Case study →

View the live system ↗

Problem: Civil engineering teams run tendering — from Notice Inviting Tenders (NIT) through estimation, compliance review and approval — as a manual, document-heavy process that is slow, error-prone and nearly impossible to track across a fiscal year.
System: SETU is a purpose-built tendering platform covering the full lifecycle: guided tender creation, document management with version control, workflow automation from NIT to approval, and a fiscal-year analytics and reporting layer. On top of that sits an AI suite — an estimate generator that produces IS 1200-compliant cost estimates, an AI chatbot for tendering and IS-code queries, and a set of free engineering calculators (earthwork, quantities, mix design) that work without login.
Stack: Next.js · Role-based access control · AI estimate generation · Vercel

02 / 05

Product / SaaS

SharpRoot — Customer Research to Cited PRDs

Live site ↗Case study →

View the live system ↗

Problem: Product teams accumulate customer interviews, support tickets and call recordings faster than anyone can read them, and turning that raw research into a defensible product requirements document is a manual synthesis job that eats a full sprint every time.
System: SharpRoot ingests interviews, tickets, call recordings and CSV exports from sources like Zoom, HubSpot, Intercom, Zendesk and Jira, extracts themes anchored to verbatim customer quotes, and generates a 7-section PRD where every claim carries an inline citation back to its source quote, file and date. Finished PRDs export as acceptance-criteria tickets directly to Jira or Linear with auto-generated story points and priority labels.
Stack: AI theme extraction · Citation-enforced generation · Jira/Linear export · MCP server

03 / 05

Global Trade

Regner Impex — Global Trade Platform

Live site ↗Case study →

View the live system ↗

Problem: A merchant export company selling across agriculture, plastics, industrial chemicals, construction materials, machinery and oil & gas equipment needed a single digital presence credible enough to open conversations with buyers in markets it had never met in person.
System: A global industrial-trade website organized around eight distinct product categories, each with its own exploration path, alongside company credibility content (About, Sustainability, Careers, Blog) and a structured inquiry flow so a buyer in any market can identify what they need and start a conversation without a call.
Stack: Next.js · Multi-category content architecture · WhatsApp + form lead capture

04 / 05

Consumer / Health Tech

Scan Label AI — Grocery Label Intelligence

Live site ↗Case study →

View the live system ↗

Problem: Shoppers stand in front of packaged food with no fast way to judge whether it's actually healthy or worth the price — nutrition labels are dense, ingredient lists are jargon-heavy, and comparing two products means reading both by hand.
System: Scan Label AI lets a shopper scan any packaged food and immediately see its health profile, price and real value under the tagline 'Eat smarter. Spend smarter.' An LLM reads the scanned label and ingredient list to flag unsafe or red-flag ingredients (additives, allergens, ultra-processed markers) in plain language, not just raw nutrition numbers. Users can compare products side-by-side, build a running basket, and dig into food-economics data that connects nutrition quality to cost, alongside a personal profile that tracks their scanning history.
Stack: Next.js · Barcode/label scanning · LLM ingredient analysis · Unsafe-ingredient recognition · Nutrition & price scoring · Vercel

05 / 05

Creative / Experiential Web

Kahin Door — Ambient Night-Journey Player

Live site ↗Case study →

View the live system ↗

Problem: Not every product is a workflow tool — this was a request for a fully custom, atmosphere-first web experience: an ambient music player themed as a late-night train journey, built to feel handcrafted rather than templated.
System: A single-page ambient experience: a looping night-train window video, a live-feeling 'window seats occupied' counter, and a themed music player streaming curated tracks (with YouTube and Spotify links) under station-based scene labels like 'Somewhere after Chhatrapati Shivaji Terminus.' Playback controls, ambient lighting toggle and a 'share journey' action are built directly into the scene rather than presented as a conventional app UI.
Stack: Custom video/audio player UI · YouTube + Spotify integration · Vercel

Most of our work is under NDA

We work with leaders and organizations around the world on projects that are often protected by NDAs. Client confidentiality comes first, and we never share names, details, or project information without explicit written approval.

View confidential case studies →