Saketh Dhulipalla @SakethDhulipal2

@AIatDoorDash San Francisco, CA Joined August 2021

Tweets

6
Followers

58
Following

79
Likes

60

DoorDash AI Research @AIatDoorDash

2 weeks ago

Agents are the new discovery layer. We've seen that across 2M+ conversations with Ask DoorDash, where people want agents to step in and carry their intent through to a fulfillable order.

Applied Compute @appliedcompute

2 weeks ago

“50% of DoorDash’s agentic restaurant orders are going to places users have never ordered from before.” @andyfang tells our CEO @ypatil125 what happens when agents become the discovery layer. If models increasingly decide what gets surfaced and bought, companies have a strong

0 3 21 11K 12

1 4 40 8K 22

View Details

Saketh Dhulipalla @SakethDhulipal2

a month ago

Our cloud-based agents platform now powers 25k+ automated code reviews each week! Here’s a great deep dive into how the @AIatDoorDash team built Flux in-house instead of relying on hosted coding agents.

DoorDash AI Research @AIatDoorDash

a month ago

Today we’re proud to announce Flux, our own cloud platform for running agents across DoorDash. Flux automated 130k engineering tasks in one month and now powers 25k+ code reviews each week. Creating the platform ourselves moved us beyond local, laptop-based agent workloads,

145 92 2K 1.1M 2K

1 2 11 682 2

View Details

Saketh Dhulipalla @SakethDhulipal2

a month ago

Thanks for having me on. Great conversation about the work happening at @AIatDoorDash and where it’s headed! Excited to finally be sharing some of the cool work that we are up to

Jamin Ball @jaminball

a month ago

Next up on First Pass: A conversation with @SakethDhulipal2 from @DoorDash With AI coding solutions exploding engineering orgs are facing a number of downstream challenges. The Doordash team has built some cool solutions to help alleviate the pain that's arisen

37 2 14 10K 6

0 0 11 741 0

View Details

Andy Fang @andyfang

2 months ago

DoorDash Air is here. Officially certified drone operator in the US. Delivering real orders, not demos "People come here because they want to solve real problems and see their work show up in the real world within days, not stuck in R&D hell watching a demo that never leaves the lab." If you're interested, apply here: careersatdoordash.com/career-areas/d…

Stanley Tang @stanleytang

2 months ago

x.com/i/article/2082…

41 27 472 971K 145

102 84 2K 780K 487

View Details

DoorDash AI Research @AIatDoorDash

2 months ago

Models out of the box have trouble reviewing DoorDash’s codebase, even when feeding them appropriate context. Single-pass AI reviewers only caught 30% of the real issues in our PRs. Instead, we built a multi-model (Sonnet 4.6 + Opus 4.8) code review agent that caught 53.6% issues at $3.91 a PR (most enterprise code review solutions perform significantly below 50% recall on DoorDash PRs). Today, we’re proud to share DashBench, our internal benchmark to compare multi-model/harness code review systems against each other. DashBench has given us confidence in maximizing the intelligence per dollar we’re spending on AI tokens to ship code safely.