Lavender S

8 + years in startups, health tech & edtech. Previously founding member @ Kyno Health. Currently on break.

[Animesh Koratana @akoratana]

We're also giving away a curated collection of 200+ Claude Code Skills our team uses daily — the workflows that made us faster engineers while building PlayerZero.

Repost and comment "100X" to get access.

[Lavender S @iamlavender20]

@akoratana 100x

[Lavender S @iamlavender20]

@samuelchenard @uiux_harshit That’s what matters. Getting things done faster allows time for more qualitative iterative loops = even more superior work

[Lavender S @iamlavender20]

Can someone explain the dumb me of how the economics of such services work for a business.
#airbound

[Naman Pushp @TheRealNamzoo]

It’s happened. We’ve just closed our seed round of $8.65 million. Led by @lachygroom, with participation from @lpolovets at @humbavc & @MohapatraHemant from @LightspeedIndia doubling down. This round has been more than a validation of our vision, it’s on the back of making it a reality.

Airbound never began as a grand vision to make 1 cent deliveries a reality. It began as the story of an engineer who questioned everything.

Why can’t this be better? Why does there need to be so much infrastructure? Why should deliveries be so inefficient?

The nice thing about engineering is that there is an objectively right and an objectively wrong answer. The nice thing about physics is that there is an optimum. It physically hurts to know what optimal looks like and go for something suboptimal.

Most engineers outgrow that feeling. The benefit of starting young is that once you get a taste for the optimum, you can’t go back. Airbound simply started as a quest to build the best drone possible. It was a project - a love letter to engineering.

We’re now a team of 50. We’ve scaled production to one drone a day, flown 10,000+ km, and improved reliability by 35x. And this was only in the past 3 months. We’re so much more than a project now.

We’ve built the best drone in the world. Now it's time for us to scale this impact and enable trillions of flights annually, and build a world where roads are optional.

Excited to have some amazing angels @balajis, @AbhayVenkatesh1, @tarunsmehta, @mohith_dzn, @dhaval_shroff, @jaredrosnerd and so many more amazing supporters of this journey. The future is going to be in the sky!

[Lavender S @iamlavender20]

Your moat. That’s that.

[tyler hogge @thogge]

Holy Moses

he’s built different.

If it’s easy, you didn’t go hard enough. Go til it hurts. Push.

I only know a few ceos who run their business this way. And guess what. They’re winning.

[Lavender S @iamlavender20]

Here it is.

[Boring_Business @BoringBiz_]

They deleted their LinkedIn account

[Lavender S @iamlavender20]

Interesting piece, considering in the age of ai just pure knowledge(fact recall) becomes useless. @reliancejio is working on its very affordable edtech product. I wish someone looks at this from their team.

[Austin Way @Austin_Way]

There are 2 types of knowledge important for understand student knowledge/mastery.

1: DOK2 knowledge, which is basic fact recall

2: DOK3 application of knowledge, FRQs, contextualization, etc.

To get to DOK3 content you must first make a student proficient in the basic facts.

I would rephrase your statement from knowledge and skills being different to mastery and knowledge being different.

(We treat a skill node as a fact or group of facts)

We teach mastery at a different time than knowledge, though a student may see both levels of content in the same session.

[Austin Way @Austin_Way]

There are 2 types of knowledge important for understand student knowledge/mastery.

1: DOK2 knowledge, which is basic fact recall

2: DOK3 application of knowledge, FRQs, contextualization, etc.

To get to DOK3 content you must first make a student proficient in the basic facts.

I would rephrase your statement from knowledge and skills being different to mastery and knowledge being different.

(We treat a skill node as a fact or group of facts)

We teach mastery at a different time than knowledge, though a student may see both levels of content in the same session.

[Lavender S @iamlavender20]

@RickelKendra @pmarca @davidsalexande3 Looks interesting. You think this is an hard read for a layman?

[Lavender S @iamlavender20]

@Copilot is the worst AI tool. Only advantage it has is the MS ecosystem data. Its prompt compliance is below standards.
Thank god I can use @comet for workarounds.

[Lavender S @iamlavender20]

@viraj_sheth I’m sure you know the oxygen guys must have some unique insights about their specific TGs(could be bodybuilders, chronic patients).

Can’t argue with the overthinking part though.

[Lavender S @iamlavender20]

@MyNykaa
Order #287510523-6335231
Feels your customer support process are outdated.
Fake delivery marked by your delivery agent. After escalation, your team made me wait forever for updates, today they closed it by giving 5% refund.

Really, Have I been scammed?

[Lavender S @iamlavender20]

3/n

How to pick good metrics and align incentives

- Find proxy metrics for long term impact, simple, imperfect
- Quantifying levers to better cross functional decisions with trade offs across teams
- Create failure metrics for edge cases, missed data

[Lavender S @iamlavender20]

Today’s podcast summed up:

1/n

Why build Data analytics team centrally Vs at Function level

- Consistent & High talent bar
- Career growth for team, better retention
- Consistency of metrics defining and methodologies

[Lavender S @iamlavender20]

3/n

Hiring and building analytics team:

- Look for curiosity as a trait
- Posing flawed ideas in front of them, observe
- See how they respond to hearing being wrong while solving case
- Check for decision making with insufficient data, chaos

[Lavender S @iamlavender20]

3/n

How to pick good metrics and align incentives

- Find proxy metrics for long term impact, simple, imperfect
- Quantifying levers to better cross functional decisions with trade offs across teams
- Create failure metrics for edge cases, missed data

[Lavender S @iamlavender20]

2/n

Balancing exploratory work(big ideas) and operational work(inbound requests) first the team

- Doing hackathons for exploratory work
- Weekly stand ups to set priorities for ops

[Lavender S @iamlavender20]

4/n

Hiring the best talent:

- Sending an unsell email at offer stage to candidates
- Touching their fears upfront (eg work life balance, environment, role potential)
- Asking politely if they still want to go ahead

[Lavender S @iamlavender20]

Summarised learnings from today’s podcast:

1/n

Getting good at writing as a Product Manager:
- Writing is clarity at scale
- Indexing towards your own taste on what’s good writing helps
- Write the way you talk

[Lavender S @iamlavender20]

3/n

Improving product sense and Decision making:

- Product sense is making good decisions with insufficient data
- Decision logs: Write others’ product decision, write your decision with rationale if you’re in their place , see the results.