Inspiration
Industrial procurement is often overlooked but essential—it's the unseen engine powering the physical world. With complex global supply chains vulnerable to disruptions, companies traditionally relied on inefficient, manual processes. The potential to streamline this critical area with agentic AI inspired us to create Hugo, an intelligent AI co-pilot designed specifically for procurement teams.
What it does
Swerve introduces Hugo, a sophisticated AI assistant tailored for Voltway, a rapidly scaling electric scooter startup. Hugo enhances procurement efficiency by predicting inventory needs, flagging potential risks, and optimizing supplier interactions. By analyzing structured and unstructured data from ERP systems, CAD files, emails, and sales forecasts, Hugo proactively identifies inventory bottlenecks, recommends strategic purchasing actions, and dynamically adapts to changing operational needs.
How we built it
We built Hugo using LangChain integrated with GPT-4o for reasoning and GPT-3.5 Turbo for efficient tool management. The system utilizes structured inventory and order data, synthesizes supplier emails and PDFs through natural language processing, and provides actionable analytics through interactive dashboards. We leveraged React and Tailwind for a sleek, user-friendly interface, visualizing complex procurement scenarios in an easily digestible format.
Challenges we ran into
Integrating disparate data sources posed significant technical hurdles. Harmonizing ERP records, CAD specifications, and emails into coherent insights required meticulous data handling and innovative processing strategies. Additionally, teaching the AI nuanced procurement logic—such as managing volatile supplier lead times and aging inventory—proved challenging but highly rewarding.
Accomplishments that we're proud of
We successfully created Hugo as a truly proactive procurement assistant. Our biggest achievement is Hugo's ability to generate real-time insights that significantly enhance decision-making. The seamless integration of AI reasoning with operational tools means Hugo doesn't just inform—he actively anticipates needs and helps procurement teams stay ahead of challenges. What's more, we built an amazing UI that is packed with information and easy to use and understand.
What we learned
This project reinforced the immense potential of AI-driven analytics in complex logistical environments. We learned how effectively AI could bridge human expertise with data-driven predictions and automation. Furthermore, developing Hugo underscored the critical importance of designing intuitive user interfaces to translate sophisticated AI outputs into practical, actionable insights.
What's next for Swerve
Looking forward, we plan to expand Hugo’s capabilities to include deeper predictive analytics, more granular supplier negotiation strategies, sentiment analysis from emails, and real-time alert systems integrated directly with messaging platforms like Slack. We envision Hugo becoming not just a procurement tool but a core operational asset, empowering companies across industries to thrive amid complexity and uncertainty.
Built With
- firebase
- graph-network
- javascript
- langchain
- langgraph
- openai
- python
- react
- slack-api
- tailwindcss


Log in or sign up for Devpost to join the conversation.