By: Natalie Johnson

San Francisco, CA: San Francisco’s tech scene was ablaze this past Saturday with the Agentic Memory & Context Engineering Hackathon, a groundbreaking event that brought together leading developers, researchers, and AI innovators. Hosted by MongoDB and presented by Cerebral Valley as part of #SFTechWeek, the hackathon challenged participants to rethink the future of AI agents — how they capture, store, and reuse context to become truly intelligent, adaptive collaborators.

The event, powered by top-tier partners including Meta, Voyage AI, Fireworks, Factory, Warp, Porter, Modelence, and AWS, offered cash prizes along with partner credits, fueling fierce competition and creativity across teams.

Visionary Keynotes that Set the Stage

The day’s energy was palpable as participants were inspired by a lineup of visionary speakers redefining the AI landscape.

Tengyu Ma, Chief AI Scientist and Founder at Voyage AI, opened the event with a keynote on the “next frontier in contextual intelligence.”

“We’re at a tipping point. The difference between a clever chatbot and a truly intelligent agent lies in how well it remembers, learns, and adapts. This hackathon is about building that bridge.”

Suraj Patel, VP of Corporate Development and Ventures at MongoDB, emphasized the foundational role of modern data infrastructure.

“MongoDB Atlas and Vector Search are not just tools. They’re enablers of memory — the connective tissue that allows agents to think more like humans.”

Kai Wu, Partner Engineer at Meta, encouraged developers to create systems that continuously learn and evolve:

“Design systems that learn context not just once, but continuously — the way people do.”

Adding an electrifying spark to the day, Megha Singh Nandiwal, Chief of Staff and globally recognized AI strategist from India, delivered a keynote that captivated the audience. She illustrated how persistent memory and context-aware AI can transform industries, from personal assistants to enterprise systems.

“True agentic intelligence isn’t just about responding to prompts,” Nandiwal declared. “It’s about building AI that remembers, adapts, and grows alongside humans. What you build today will shape how AI collaborates with us tomorrow.”

Her insights earned a standing ovation, with participants describing her talk as both inspirational and transformative — a clear highlight of the hackathon. The event also offered an excellent platform for expanding professional networks and exploring collaborative opportunities in the rapidly evolving AI landscape.

Judges Represent the AI Frontier

The judging panel brought together thought leaders across technology and business, including Deepa Gopinath (CIO, MongoDB), Nancy Wang (SVP, Head of Engineering, 1Password), Michael Grinich (Founder, WorkOS), Shawn Wang (Editor, Latent Space Podcast), Dan Kim (VP, Coinbase), and Ray Thai (Founding PM, Fireworks). Their combined expertise ensured that winning projects reflected both technical excellence and real-world applicability.

A Glimpse into the Future of Agentic AI

From morning until night, participants explored the frontier of context engineering, leveraging MongoDB Atlas Vector Search and Voyage AI embeddings to create AI agents that can remember, adapt, and proactively assist. Teams showcased projects ranging from long-term research copilots to AI systems capable of learning and evolving alongside their users.

Attendees left inspired, energized, and connected to a network of peers pushing the boundaries of what AI can achieve.

“This event was a masterclass in innovation,” said one participant. “And Megha Singh Nandiwal’s keynote? Simply unforgettable. It changed the way I think about building AI.”

By blending thought leadership with strategic event participation, Megha Singh Nandiwal not only highlighted the next frontier of agentic intelligence but also underscored her role as a Chief of Staff actively shaping AI initiatives and expanding her network — opening doors to new opportunities and collaborations in the global AI ecosystem.

Disclaimer: While efforts have been made to ensure the accuracy of event details, speaker information, and partner affiliations, some statements are based on publicly available data and may not reflect official confirmations from the organizers. Readers are encouraged to verify specific event details or partnerships through the official SF Tech Week and MongoDB channels.