Co-founder & CEO @BanzaOfficial
Dubai, United Arab Emirates
Joined January 2013
- Tweets1.9K
- Following3.9K
- Followers1.7K
- Likes13K
The smartest part of Walmart's AI agent rules is the profile. Your agent gets its own budget and its own purchase history, and if it buys the wrong thing the return still works. That's how most of us got our first bank card. A small limit, then more trust once it proves itself.
Most of the AI at Dubai AI Festival was built for companies.
Very little was built for the person on the other side.
Companies are getting their agents. I think people need one too, one that knows them and is on their side. That's what we're building at Banza.
I’m at Dubai AI Festival today with @BanzaOfficial
I’ll be at our booth, S88. Come by and I’ll show you your AI Twin first hand: it learns what you buy, then finds you better deals on it.
If you’re around, come say hi !
It’s been a long time my friends. Now that actually this new algo changes works, I have a lot to tell you. Many updates on work , on personal life … I will be sharing more from now on because I know at least friends can see it !
I’m really excited to be joining the @arts_dao Circle 🤝
Big respect to @Rahim_mahtab and the team for building something that actually feels like a proper home for builders and quality people, real connections, and that IRL clubhouse energy we’ve been missing.
Looking forward to contributing, learning, and creating alongside this group.
JOINING THE CIRCLE: @Mehdi_
Co-Founder & CEO of Banza, an AI-native platform where users build and own their personal AI Twins. Previously led BD for PrivadoID at Polygon Labs and founded Blocklogica Labs, a Web3 venture studio in the UAE.
Welcome Mehdi 🤝
Mehdi retweeted
🚨 HIRING ALERT: Banza Engineering 🚨
We are hiring 2 engineers to materially shape the next year of our product.
Both roles are Remote-first (IST); Bengaluru-based is a plus. First hire in seat within 30 days.
The Roles:
💼 Applied AI Engineer
We need an engineer with deep practical experience building modern AI systems, whose technical judgment will raise our team's ceiling and compress weeks of trial-and-error into sharp, well-reasoned decisions.
The Work:
-Ship end-to-end AI features from idea to production.
-Set the bar for reliability, evaluation, and quality across AI systems.
-Balance latency, cost, and quality across the inference stack.
-Elevate team thinking on agentic systems, retrieval, and model behavior.
-Balance where deterministic systems end and probabilistic ones begin.
What We’re Looking For:
- 3+ years in engineering (5+ preferred); 2+ years using AI-native systems.
-Last 24 months in a lean startup or building solo with full ownership.
-Shipped production AI features to real users in real-time (no demos).
-Built non-trivial RAG systems; strong views on retrieval and grounding.
-Comfort with evals, benchmarking, and reliable probabilistic systems.
-Fluent with AI coding tools, agentic workflows, and AI-native dev.
Counts in Your Favor:
-Building side projects, OSS contributions, or weekend experiments outside of work.
-Hackathon wins or public writing/talks on agentic systems, retrieval, or LLM evaluation.
-Experience shipping consumer-facing AI features (vs. internal tooling or B2B).
💼 Product Engineer
We are looking for a mid-to-senior full-stack engineer with a mobile-first center of gravity who treats shipping as a craft. This is a bandwidth and resilience hire to expand what we can ship per sprint.
The Work:
-Own end-to-end consumer mobile features from interaction detail through --backend services.
-Contribute meaningfully to backend systems and move fluidly across the stack.
-Raise the bar on shipping velocity, code quality, and product polish.
What We’re Looking For:
- 3+ years in engineering (5+ preferred); 2+ years using AI-native systems.
-Last 24 months in a lean startup or building solo with full ownership.
-Shipped at least one consumer mobile app to production.
-Flutter experience strongly preferred (advances you over other candidates).
-Comfortable in Node.js backend territory.
-Fluent with AI coding tools, agentic workflows, and AI-native dev.
Counts in Your Favor:
-A genuine eye for craft, polish, and the small details that separate "shipped" from "loved."
-Familiarity with JavaScript runtimes and browser internals (useful for plugin/protocol work).
-Building side projects, OSS contributions, or winning hackathons.
Logistics & Application:
Compensation: Up to 30LPA + meaningful equity (discussed privately based on seniority).
How to Apply: Download the app, comprehend Banza's AI Twin functionalities, and complete the application process here: forms.gle/rvjfmfpdyqZxbDXd9
After spending years working across different markets, ecosystems, and teams, one thing became increasingly clear to me:
AI still doesn’t truly understand people.
Most systems today operate on fragments:
a prompt, a few clicks, short term context.
But people are far more than isolated interactions.
We spend years generating patterns through our habits, preferences, routines, behaviors, and decisions across platforms, yet almost none of that context actually works in our favor.
That realization is what led us to build @BanzaOfficial .
Today, I’m grateful to share that Banza App has raised a $1M pre seed round led by @YourCampusFund , with participation from @avax and an incredible group of angel investors.
At @BanzaOfficial , we’re building AI Twins powered by user owned data:
personal intelligence systems that deeply understand individuals and proactively help them make better decisions across platforms.
I believe the next generation of AI products will not be defined only by model intelligence.
They’ll be defined by how well they understand the individual behind the interaction.
200k+ AI Twins have already been built.
We've raised $1 Million in Pre-Seed funding, led by @YourCampusFund, to define a new standard.
Everyone got comfortable ignoring the problem that AI got smarter, but never for you specifically.
We built the Banza AI Twin. Yours. Trained on your actual behavior, 100% working in your interest.
Meet yours. banza.xyz/app
Mehdi retweeted
About 400,000 incredible Iranian men and women call the UAE home. Their children grow up here. Their families build businesses here. Their future is here. When aggression threatens, it does not ask for passports. One Iranian was injured and the UAE defended them as it defends more than 200 nationalities who live side by side on this land. This is what leadership looks like. This is what peace protects. The UAE is strong. The UAE is safe.
Finally got my @freedomfactory and i have to be honest its much cooler than what i expected.
Mehdi retweeted
$27 trillion settled.
Stablecoins processed more value last year than Visa's entire annual payments volume.
J.P. Morgan moves $2B+ daily on Ethereum rails.
New report from Etherealize, Nethermind, and L2BEAT on the institutional Layer 2 landscape.
Still need more reasons?
The ecosystem since 2022:
- Networks: 27 → 154
- Throughput: 47 → 1,500
- TPS TVL: $35B+ across leading L2s
Even more reasons:
- J.P. Morgan piloting tokenized deposits on Base
- Société Générale issued stablecoins under MiCA
- Visa prototyping ZK proofs for private asset transfers
-BlackRock and Franklin Templeton running tokenized funds
Moooore Reasons you asked for:
Regulatory clarity is emerging.
- MiCA is live in the EU.
- The GENIUS Act provides a US stablecoin framework. - Singapore and Hong Kong are approving tokenized products.
For institutions, L2s solve a core tension: public blockchain security with private network control.
Custom compliance, transaction confidentiality, guaranteed service levels.
All settled to Ethereum's neutral ledger.
First we have seen this with aave and now metamask . It’s amazing to see major products fixing one of the biggest issues “trust” . I assume next ones will be smart contract insurances . I hope soon we cover the whole process
Great call by the MegaEth team. It is incredibly rare in this space to see someone return half a billion simply because they felt they didn’t execute it the way they intended. Another clean W for them, and it sets a tone that people will remember.
