Prharsha

Hey guys!

We've been improving Atlas, our local knowledge base for the agents.

Over the past few days we've been running a series of retrieval experiments to improve how Atlas understands natural language when exploring large codebases. The early results are promising, with significantly more relevant entry points on semantic queries while preserving keyword search.

We're now benchmarking everything before sharing the full findings later this week.

Thank you to everyone who followed Tempest during our launch

We may not have had the Product Hunt launch we hoped for, but we did gain 18 followers, and we're genuinely grateful for every one of you.

We're just getting started.

Going forward, we'll share weekly updates here on Product Hunt so you can follow our progress, see what we're building, and get a behind-the-scenes look at Tempest as it evolves.

We're building Tempest because we think agentic engineering is heading in the wrong direction.

Today, most tools compete by running more agents. More agents sounds impressive, but it also means more duplicated context, more token usage, and more coordination problems.

We think the real challenge is different.

How do you help every agent understand the codebase without making each one rediscover it?