Personal Product · 2026
In progressEliminating context loss in product teams
Unify is a context router for product teams. You capture once - screenshot, URL, or note - and the full context is preserved and delivered where it is needed. No digging in Slack. No "where did we decide that?" pings. Just context that shows up on time.
Role
Founder, Product Designer, Builder
Stage
Early MVP (v0.1.0)
Focus
Context routing + intent preservation
Status
Live and evolving
View live MVP
still in the workshop · shipping in public
The problem: work remembers badly
Product teams make decisions for good reasons. Then meetings end, tabs close, and people context-switch. Three days later the reason is gone, and the team rebuilds the same context from scratch.
Unify starts from one belief: summaries are useful, but raw captures are truth. If the source survives, intent survives.
I built this because I was tired of teams re-litigating decisions they already made. Memory should be a feature, not a weekly fire drill.
— Gautham Ravikiran
competitor.com/pricing
$49/month
Unify
U
U
Figma
FileEditView
Layers
Page 1
Hero
Nav
Cards
Footer
Context from Unify
Pricing Research
Competitor charges $49/mo. Users show price sensitivity above $50.
Key Decision
Team agreed on $45-50 range based on competitive analysis.
Source: Chrome capture - 2h ago
Capture
Route
Surface
How it works
Capture, structure, retrieve, inject. Same context, different moments in the workflow.
01 · Capture
pinterest.com/explore/ui-grids
P
Search for ideas
HomeCreate
Capture to Unify
Screenshot
URL
Text
Current page preview
Why does this matter? (optional)
e.g., Reference for our grid layout
Capture
Pinterest Grid Layout
screenshotpinterest.com/explore/ui-grids
Your note
"We need the grids to be like this"
Summary
Pinterest-style masonry grid layout with image cards of varying heights arranged in responsive columns. Rounded thumbnails with caption text below each card.
Key Points
- Masonry layout with variable-height cards
- 3-column responsive grid structure
- Rounded image thumbnails with captions
grid-layoutmasonrypinterestui-reference
Capture is the trigger. From the browser extension you save a screenshot, URL, or selected text and add an optional note. Full context is stored—nothing summarized away. That capture is the start of the route so it can reach the right person or tool.
02 · Search and ask
LeadCapture.tsx
1function LeadCapture() {
2 const [email, setEmail] =
3 useState('')
4
5 return (
6 <div className="form">
7 <h2>
8 Start your free trial
9 </h2>
10 <input
11 placeholder="Email"
12 value={email}
13 />
14 {/* TODO: Visual design? */}
15 <button>
16 Get Started
17 </button>
18 </div>
19 )
20}
Unify Context
Ask about your context...
SC
Sarah Chen
Designer
Design Notes
CTA #2563EB • 16px pad, 12px radius • input 44px
lead-capturedesign-specvisual
Search the capture list or ask in plain language in the web app. You get the actual context back—snippets and sources—so no "Hey, do we have that competitor research?" in Slack. We're building toward context that finds you in the tools you work in (Figma, VS Code).
03 · Inject
Unify Context BlockLead capture design specs • 3 sources-> ChatGPT
ChatGPT
Unify Context
Lead Capture Design Specs
CTA: #2563EB • Pad: 16px • Radius: 12px
ChatGPT
Figma
Code Editor
Figma
Context panel with design specs injected into the design workspace.
Code Editor
// Unify Context: Lead Capture Design Specs
// CTA: #2563EB | Pad: 16px | R: 12px
export function LeadCapture() {
return <button className="bg-[#2563EB] p-4 rounded-xl" />
}
ChatGPT
Figma
Code Editor
Package context for where it's needed: select captures, add a short goal, get a context block. Paste into Claude, Cursor, ChatGPT, or a handoff doc—or have it surface in a teammate's workspace. Full context, no loss.
Design decisions that shaped the MVP
- Capture in under 5 seconds
- If capture feels like admin work, nobody uses it. The extension keeps capture dead simple: screenshot, URL, or text with an optional note.
- Conversation over query syntax
- People do not remember exact tags. They remember questions. So retrieval is built around natural language prompts and source-backed results.
- Context block, not black box
- Injection is transparent. You see source and key points before sending context into AI chats or handoff docs.
- Respectful by default
- Unify only captures what you explicitly choose. It behaves like a quiet teammate, not a surveillance bot.
Tech stack
- Frontend: Next.js, React, TypeScript, Tailwind CSS
- Capture: Chrome/Edge browser extension workflow
- AI layer: OCR, summarization, semantic retrieval
- Delivery: context blocks for AI tools and handoffs
This is early, but useful. The current MVP proves the core loop: capture once, retrieve fast, inject with intent intact.
What's coming next
The MVP is live; the roadmap is about context finding you instead of you finding it. Here's what I'm building toward.
- Context in the tools you use — Figma and VS Code integrations so relevant captures surface when you open a file or project, without leaving the app.
- Context that reaches teammates — Push context into a teammate's workspace or handoff doc so the "why" travels with the work.
- Smarter linking and search — Related captures and better semantic search so you can ask in plain language and get the right slice of context back.
If that sounds useful, the live MVP is the best place to see where it's at today — and you can sign up there to hear when new pieces ship.
View live MVP and early accessNext project
Designing Accessible VR Therapy for Aging Patients
A complete redesign of a VR therapy app for one-hand navigation and independent use by patients recovering from thoracic surgery.