Vendra set out to solve the chaos in hardware vendor management. When I joined as the first team member, there was no product—just an idea. To help hardware teams manage their projects and vendor communication with ease.
I visualized and built the entire platform from scratch, experimenting with workflows and refining the vision through multiple iterations.
Starting with an MVP for centralized vendor communication, I quickly identified challenges with user adoption and designed the pivot — to layer AI-driven insights over existing workflows — providing a search experience that is the single source of truth. Thus we created a powerful yet intuitive solution tailored for hardware teams.
Over 18 months, Vendra evolved into a polished SaaS platform, securing multiple paying customers and $500K in pre-seed Y combinator funding. My role encompassed everything from product direction and design systems to strategic pivots, reflecting my ability to transform an idea into a market-ready solution.
SHOWCASE
The platform in a Nutshell
0.1
Athlete Workout
0.2
Athlete and Coach Analytics
0.3
Discovery and Teams
0.4
Coach <> Athlete Communication
THE PROBLEM
Fragmented Communication with Vendors
Hardware teams are stalling across the industry
The hardware industry relies heavily on disparate communication channels during projects with vendors— from emails to chat apps like Slack and WeChat — resulting in fragmented information and project delays.
30% of a manager’s time is wasted due to miscommunication and lack of centralized information.
4-6week delays reported by Hardware programs on avg due to poor information and communication management.
90% of engineers reported desire for faster access to up-to-date data
INITIAL OFFERING
A Central Hub for Seamless Vendor Collaboration?
Designing the MVP
I led the ideation process to create a platform that consolidates a dedicated vendor communication thread, a centralized vendor library, and a project management tool.
We created a single source of truth for hardware teams - eliminating the need for scattered updates.
Our MVP showed promise, securing a spot in Y Combinator with $500k pre-seed funding based on its potential to transform vendor management for hardware teams.
Dedicated task space for collaboration
Instead of distributed communication channels, every task has a dedicated page for the exchange of information between entire hardware teams and vendors. The singular task thread captures all messaging needs with and automatic translation of language for international teams - eliminating the use of multiple conversation platforms.
0.1
Task Page - Comprehensive collaboration space
Systematic management of project timelines
The project page is a consolidation of all tasks that the teams are working on. Hardware teams can easily track the progress of each task based on priority and be on the same page as vendors when it comes to progress tracking.
0.2
Project Page - Tracking Progress
Central Vendor Library
Hardware teams working with multiple vendors are overwhelmed by the number of different simultaneous projects. The vendor library gives them a robust directory to manage several different projects with the same or different vendors.
0.3
Vendor Library - Seamless management
Organization Dashboard
A broad overview of all concurrent projects and tasks. It set the context for the hardware teams for a quick access to recent, ongoing or blocked tasks.
0.4
Dashboard - Quick Access
ADOPTION CHALLENGE
Addressing the pitfalls of over-engineering
Why weren't people using it?
Not solving the root of the problem
We got a bit ahead of ourselves with what we believed to be the real use cases. But as the design process often reveals, we needed more feedback from the users.
Testing with real users, our first customers - revealed resistance from vendors who preferred their existing communication tools, leading to onboarding friction.
The breakthrough insights!
Hardware teams needed project management support more than vendors
A single team has multiple vendors of just one project or even for a single part. A vendor has a single point of contact for their project - the company they are working with. Therefore hardware teams have a much more convoluted process for managing just one project.
Data retrieval for unstructured data is harder than just communication
Even if all projects and communication were on the same platform, it still takes a lot of time to retrieve information and stay on top of your game. Going through different communication channels with different vendors on the same platform to capture valuable information still remains an issue.
Users were trying to force AI tools that didn't fit their needs
The use of tools like Chat GPT could help only to an extent at deciphering their needs as it lacked context. Providing context manually for every project and conversation is not just inefficient but practically impossible.
So... what did we do about it?
Recalibrate and Pivot
It's okay to learn from your mistakes. It was more important to understand what the users really wanted.
In the latter half of the project evolution - I designed the pivot to leverage AI integration with existing channels like Gmail, Slack, and WeChat, turning the concept into reality within a short turnaround.
STRATEGIC PIVOT
Testing a new direction - Leveraging Generative AI
The path of least resistance - a personalized approach
Automatic extraction of data and context
I believed the easiest would be to simply add Vendra to emails or group chats with vendors and suppliers and Vendra AI extract key information like project details, shared files, and more.
Testing a familiar Conversational AI interface
A context-aware AI assistant simplifying current workflow using RAG (Retrieval Augmented Generation)
The approach was to help users manage projects, files, and every other nuanced interaction with vendors, suppliers, and manufacturers without having to spend countless hours on deciphering communications.
1.1
Chat interface - v1
We had our Aha! moment guys...
This really clicked with hardware mangers.
The Beta version of the context aware AI conversational assistant was an immediate success. This helped us find confidence in the solution and we started pitching heavily to clients.
Beyond conversations - Discovering new needs
Managers needed to stay on top of all the data that was exchanged, created or augmented. Apart from retrieving conversations, we could retrieve unstructured data from all platforms used by Hardware teaming, from CAD to JIRA!
Vendra Chat - Talk to Vendra AI and ask anything about your projects. Literally.
Once the user connects all the different apps that their enterprise uses, Vendra can index any number of files, documents and conversation media and retrieve answers for you with RAG.
Using RAG prevents hallucinations (which is often an issue with AI assistants) and provides a seamless learning experience for the user.
2.1
Chat results UI
Customizable retrieval scope for sourcing more accurate context
From user behavior we learned that not all chat require indexing every relevant document. An expert user can recognize more important documents in context of the conversation.
To help them - we created a smart summarizer that can give you a quick overview before focussed chat with updated referencing through RAG.
2.2
Chat components - zoom in
Vendra Search - Fast access to generic questions
Not all questions require a chat and a deep dive, some use-cases are covered by a simple search and retrieve function that interlinks platforms to save time and clicks.
2.3
Vendra Quick Search results interface
Built around a hardware manager's ideal workflow. With Bots.
Communication with internal and external teams happen of different platforms like Slack, MS Teams and Google Workspace.To have vendra integrate with such platforms made the most sense.
2.4
Slack Bot Integration
REFLECTION
Single handedly leading design efforts at an AI startup calls for serious growth
Silent contributions often get overlooked -- Yes I'm talking Design Systems
Not all startups need a fully formed design system. Tailwind's CSS might be more than enough to get the product out of the door.
But does it add character? No.
To keep things lean, I created key components and color schemes that could be combined seamlessly with Tailwind CSS design system. This helped rapid development and removed design debt.
0.1
Task Page - Comprehensive collaboration space
Key learnings
In the age of AI, meeting business needs is key
While we are in this rapidly evolving AI paradigm, companies are quick to move on available Gen AI capabilities. What was important for me was to be able to navigate this space and acquire a whitespace that makes the product viable.
Differentiate or Die
During our MVP stage, our product didn't serve a specialized use case. Making our product's key features indistinguishable from competitors like Asana and Jira. But now - we serve unique needs of our key users providing a smarter way to build and ship hardware faster - by saving hours for hardware managers.
Looking for some impact metrics? Stay tuned - it's coming soon!
Apart from raising $500k and getting backed by Y-Combinator - we just made our product live to 21 hardware teams.
For me, that's quite impactful as a designer, but you're gonna see some numbers soon! Brb.
