Command your own software factory.
San Francisco
Joined March 2025
- Tweets211
- Following32
- Followers1.3K
- Likes264
Macroscope retweeted
your eng team burns hours reviewing PRs an agent could have approved. macroscope fixes this. "Approvability" is our agent that auto-approves PRs that are safe to merge.
today we're launching a dashboard that shows how Approvability is performing, what it approved and why, so you can confidently make it more permissive over time and understand the impact.
introducing the Approvability Dashboard: see what macroscope's approvability agent is approving, what it isn't, and how your policy changes move the numbers.
approvability works out of the box with sensible defaults. want it more permissive or tuned to your team's standards? add your own rules in "approvability.md", then watch how those changes play out on your real PRs:
✅ approval rate: the share of PRs macroscope approved, plus a ceiling for what it would have approved
🛡️ eligibility: how often the agent judged a change safe to approve without a human
🚦 verdict mix: approved, would approve, needs human approval, skipped
📋 every PR and its verdict, searchable by repo and time range, and exportable to CSV
find it under code review → approvability in the macroscope app.
Macroscope retweeted
@Macroscope runs its go-to-market on a business world model built with Lightfield.
They ingest customer touchpoints across marketing interactions, product use, and sales into a single platform. This gives agents full context on entire customer lifecycle. Agentic workflows that used to take months to build in other systems can be designed in hours using natural language.
Read how Macroscope is building its GTM system.
lightfield.app/customers/mac…
Macroscope retweeted
@cars24india is one of India's largest auto-tech companies: 300M+ users, $1B+ in revenue, 400 engineers.
with Macroscope, they saw up to 3x more code reaching production in 3 months. 🧵
Macroscope retweeted
“A single engineer might be capable of directing dozens of productive agents, yet in practice can only keep a handful moving at once.”
Exciting early preview from @kayvz and the @Macroscope team, making it easier for engineers to direct fleets of coding agents in the cloud and for work to stay in motion.
Macroscope retweeted
Murmur has significantly improved my productivity and changed the way I work. Being able to trust agents to self-drive from implementation through verification and fixes has been a game changer.
Super excited for this! Happy spawning 🚀
Macroscope retweeted
Murmur cloud saved my laptop. 🐦
When model intelligence is no longer the bottleneck, how you run your agents is.
That’s the problem Murmur exists to solve and it’s also how we ship code at Macroscope now.
Murmur is now in early preview. Excited to bring it to more teams!
Macroscope retweeted
I live and breathe MURMUR.
Here’s how I use and abuse this feature to create a metrics dashboard (internal tools only):
Macroscope retweeted
Been using Murmur internally every day, and it’s genuinely changed how I write software.
The biggest shift isn’t that agents can write more code — it’s that I can keep a whole fleet of them moving in parallel while staying focused on the decisions that actually need me.
One of my favorite parts: I can just hand Murmur a PR and have it spin up a VM with the resources it needs, run the tests there, and post the results back to the PR. No more sacrificing my localhost to some massive test suite 🥹
Very excited to finally share what we’ve been building.
Macroscope retweeted
Working with Murmur is magical 🪄
It strikes the perfect balance between letting your agents rip and having enough visibility over what they are doing. It runs in the cloud, gives you snapshot artifacts of what it has done, and it seamlessly integrates with @Macroscope code review.
Macroscope retweeted
the @FireworksAI_HQ team have been fantastic partners. they provide near-real-time support and are obsessive about providing customers a fantastic experience. even in the rare case where we've ran into hiccups or bugs (things happen!), they were so impressively fast at resolving and getting us back on the right track that it increased our love and gratitude for their service-- this says a lot about their ethos towards customers.
we'll have more to share soon on what we've built on their training platform, and the results :)
thank you @chahvivi @lqiao @the_bunny_chen for all of your support and for building something super valuable for the AI ecosystem
every conversation with @Rob_Bishop has been a insane learning experience. @Macroscope was one of the earliest power user of the training platform and its been a fun journey to keep up with the amazing work they are doing. Deeply technical, relentless on quality bar, always pushing our training infra to the limits, and fun to work with. thank you @kayvz and @Rob_Bishop for helping make the product better with your feedback.
Macroscope retweeted
A noisy code reviewer gets ignored. Flag too many non-issues and engineers tune it out.
@Macroscope got the control they needed with the Fireworks Training API to train a model using their own advantages and loss to improve the signal-to-noise ratio.
Build your own frontier: fireworks.ai/training
Macroscope retweeted
👀 @macroscope is now #1 on the latest online Martian code review benchmark on F1 score (which balances Recall and Precision)
Macroscope retweeted
"Most B2B tools are atrocious."
@kayvz (co-founder of @macroscope) answers @lennysan's question about building B2B after a career in consumer products:
"My first company actually was a B2B company, selling software to universities."
"After that experience I was like, B2B SaaS, never again."
"Maybe next to government, it's the slowest possible sales cycle."
"I actually think there's tremendous alpha bringing consumer instincts to B2B tools.
They're so poorly designed.
They're so focused on being functional that they lack soul, they lack joy."
"They're just not built with love for end users in mind, because you're often selling to a person that's different than the end user."
Macroscope retweeted
code review performance is an enormously complex series of trade-offs between recall (bug detection), precision, latency and cost.
we work very hard to find the perfect balance and abstract all the complexity away for our customers so that you just get exceptional code review without hemorrhaging money.
Sadly, despite a sweet analytics dashboard on Greptile, I can't find any wall time data for reviews anywhere.
But looks like 6 to 7 minutes in situations where like Macroscope and CodeRabbit fished super fast. Like 1m for Macroscope. Sadly CodeRabbit doesn't report the time in the CI summary, and despite having an entire page of "Time Metrics" not one of them is how long it takes to do a review (but its done long before Greptile). You both are all about like Time to Merge metrics... though that sort of data is of little use in my experience.
What kills me is the iteration time... AI finds lots of random things, and is never exhaustive... so you get things to fix, you fix them, then it re-reviews and finds more things, round-and-round until its clean. So the wall-clock issue I am having is the slowest thing in the chain (since everything is in parallel), and that is Greptile these days. It use to be fast, so something changed.
Macroscope retweeted
@kayvz is someone I’ve long admired, and what he and @Macroscope are building is truly incredible. We’re honored to be hosting this event with the Macroscope team! Don’t forget to RSVP!
We’re hosting an event with @Google on July 14 that I’m really excited about: In Good Company. Founders and engineering/product leaders will demo the internal AI tools and products their teams rely on every day. Stuff that’s usually unseen publicly.
If you’re a founder, eng/product leader, or engineer, come join us @ Google SF to learn, mingle and snack! luma.com/ao9tgos2
Macroscope retweeted
this is shaping up to be really great. come thru!
fireside talks with: the founders/ceos of @maticrobots, @frecfinance, @descript, and @WeekendHQ
demos by leaders at: @pydantic, worldcoin, @lightfld
joined by eng and product leaders from: OpenAI, Robinhood, Brain Co, Luma AI, Console, Air Ops, Doordash, Snowflake, Figma, Persona, Salesforce, Retool, Granola, Box, Ford Motors and many more.
WE ARE ALSO LOOKING FOR A MUSICIAN TO MAKE THE B2B SAAS BOONDOGGLE VIBES IMMACULATE PLZ HOLLER IF YOU ARE A TALENTED HARPIST, GUITARIST OR SOMETHING
We’re hosting an event with @Google on July 14 that I’m really excited about: In Good Company. Founders and engineering/product leaders will demo the internal AI tools and products their teams rely on every day. Stuff that’s usually unseen publicly.
If you’re a founder, eng/product leader, or engineer, come join us @ Google SF to learn, mingle and snack! luma.com/ao9tgos2