← All stories
● Covered by 1 source · 1 reportLow impact1 neutral

Lemonflow uses Gemma and TPUs to build real-time AI agents for EV charging support

🔄 Updated 10d ago
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • 15-25% of EV charging attempts in Europe fail.
  • Lemonflow's AI agents assist drivers every 90 seconds.
  • Agents use Gemma and TPUs for low-latency, real-time support.
  • The system supports over 90 languages across multiple communication channels.

Addressing EV Charging Failures

Lemonflow, founded by Dr. Lukas Fieber and Konstantin Huneke, developed AI agents to tackle the high failure rate of electric vehicle (EV) charging in Europe. Between 15% and 25% of EV charging attempts in Europe do not complete successfully, often leaving drivers stranded or waiting in long call center queues. The company's AI agents provide support to EV drivers and monitor charging networks.

Real-time Conversational AI Requirements

Building real-time voice support for EV charging presented several technical challenges. Strict latency thresholds were necessary, with latencies above 800 milliseconds risking conversational interruptions. Initial reliance on managed LLM APIs led to latency spikes during peak hours, reaching over 1.7 seconds, and HTTP 429 rate-limit errors. The agents' design, using comprehensive prompts and dynamic tool calls for live charger data, created intense, bursty compute demands difficult to guarantee on shared capacity.

Solution with Gemma and TPUs

To overcome these challenges, lemonflow implemented AI agents using Gemma and TPUs. This infrastructure allows the agents to provide immediate assistance, answering drivers in over 90 languages. The system is designed to resolve real-time charging issues across various communication channels, including phone, chat, email, and WhatsApp, handling a fragmented ecosystem of approximately 30 distinct charge point operators.

Impact on EV Charging Experience

The deployment of these AI agents aims to improve the EV charging experience by reducing failed sessions and providing instant support. By automating the resolution of common charging problems, lemonflow seeks to contribute to the broader adoption and reliability of electric vehicles in Europe's growing market.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~4 min · 3 stories · Nov 21

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Reporting from

Lemonflow developed AI agents for EV charging support that utilize Gemma and TPUs to address high failure rates in EV charging. These agents provide real-time assistance to drivers in over 90 languages, aiming to resolve issues like stalled charging points and slow call center queues.