Building Soda to monitor from data to decision

I love talking about: machine learning, data quality, decision making, post deployment data science, performance estimation, and bunch of other stuff. 

Always open to have a chat 🙂

Hakim Elakhrass

I'm the founder of Octolens, a brand monitoring tool built for developer and AI companies. With Octolens, you never miss critical conversations about your brand, competitors, or customer pain points. We surface high-signal posts so you can engage, connect, and grow.

Over 1,000 founders and growth teams use Octolens to stay close to the conversation.

Before Octolens, I was Chief of Staff at crowd.dev (acquired by the Linux Foundation), where I worked on DevTool growth and community-led strategies. I also co-founded SaveSpace (Techstars '22).

Earlier, I worked at brand consultancy Prophet, where I helped businesses refine and protect their brand identity.

At Octolens, we help teams monitor their brand online—because what people say when you're not in the room matters.

Charlotte Schmitt

One thing I didn't expect to be so hard when building Octolens: explaining our killer feature.

I just had a demo call with a founder, he explained:

"We signed up to Octolens at the same time as [competitor] and they found these mentions you didn't" 

They are a health tech brand and their name is used regularly in a completely different context. 

They continued: "We'll need to tweak the other tool to stop getting these irrelevant mentions. But why didn't Octolens find them?"

Well, we did find them. We just didn't alert you.

Before we ping you, we check whether the post actually relates to your business, not just whether your brand name appears in it.

The other tool sent the noise to you, we filtered it out automatically.

Maybe we've made this feature too invisible? If you're an Octolens user and have thoughts on how we should surface this better, I'm all ears.

Founder at Octolens | Brand monitoring for AI & developer companies

"great marketing, but it's BS right?"

That was a DM I got after posting our origin story last time. 

My reply: "Nope. Worst 6 or so weeks of my life." 

Because two years ago, I really did spend hours, and hours, and HOURS, doing manual searches and sending emails with mentions to our first customers.

Not fun. But those 6 weeks were some of the most valuable for the company.

1) I validated that people would actually pay without a single line of code, just a Stripe link and me as the "AI social listening product."

2) I learned very fast that the problem isn't tracking posts. It's getting rid of the noise.

One of our first customers was Juicebox, an AI recruiting tool. Also a children's snack. Also used in a other contexts. 

Another customer's brand name was a common word in Japanese. 

Another, a piece of machinery.

And that's just from a dozen or so starting customers.

Weeks of manual sorting gave me examples and edge cases I needed to build an AI assessment that was as good as a human going through it for B2B SaaS. 

While other tools relied on long lists of inclusion and exclusion terms, we used company context and everything I'd learned the hard way.

Dogfooding your own product is great advice.

Why not try being your own product?

taplio

The Battle

Hakim Elakhrass

Charlotte Schmitt

Followers

Posts

Engagements

Likes

Performance ScoreCharlotte Schmitt's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Hakim Elakhrass

Charlotte Schmitt

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Charlotte Schmitt's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreCharlotte Schmitt's LinkedIn Performance