Track your favourite FinX accounts. The moment they name a new stock, we send you a fact-checked deep dive; thesis, financials, risks. Not investment advice.
Joined January 2026
- Tweets89
- Following35
- Followers141
- Likes324
Serenity flagged $SIVE below 5 sek, it touched 110.
Today the market in panicking with $SIVE touching 49, and missing the actual signal here.
June 30: a directed share issue of ~$73m to institutional investors at 57 SEK. So OVER the current price.
Q4 2026: series production begins on lasers and potical amplifiers for a major customer's LIDAR platform.
Next few quarters: NASDAQ listing finalized (according to CEO Vickram Vathulya recent statement).
2H 2027: the CPO ramp. Sivers' lasers are in qualification now, targeting production-ready status in 2027, timed exactly to the customer volume ramps their partners are guiding for 2028.
For more, link in bio. We log every ticker Serenity mentions, and write deep dives.
The public market finally has a humanoid setup it can actually buy.
Tesla Optimus is trapped inside $TSLA.
Figure is private.
Unitree is private / China.
Boston Dynamics is inside Hyundai.
Agility could become the default U.S. humanoid robotics ticker before the category is obvious.
Digit already has real deployments, 65,000+ operating hours, 9 customer facilities, $300M+ Digit v5 orders, and backing from $NVDA, Amazon, Foxconn and SoftBank.
Short thesis:
Humanoids move from demo videos to enterprise deployment, and $CCXI becomes the first clean public proxy.
The market may be getting its first real U.S. humanoid pure play.
Tesla Optimus is trapped inside $TSLA.
Figure is private.
Unitree is China/private.
Boston Dynamics is inside Hyundai.
While Agility is trying to become the public ticker.
Agility has:
Digit already deployed in industrial environments
65,000+ operating hours
9 customer facilities
$300M+ Digit v5 orders, subject to milestones
NVIDIA / Amazon / SoftBank / Foxconn ecosystem
Foxconn-led PIPE
RoboFab capacity path toward 10,000+ units/year
The bull case is simple:
Digit is built for boring industrial work, which is probably where humanoids commercialize first.
Warehouses.
Factories.
Tote movement.
Repetitive labor.
Human-designed spaces.
Then the moat compounds through deployment data.
Every site gives more operating hours, more safety validation, more workflow learning, and more enterprise trust.
Add NVIDIA for physical AI/safety compute, Foxconn for manufacturing scale, Amazon/logistics demand, and SoftBank-style category capital.
$CCXI is now the ticker to watch.
Serenity just opened a brand-new position. Is this the next $SIVE? The next $SNDK?
$CCXI (changing to $AGLT on IPO this September). Agility Robotics. The first pure-play humanoid robot company to go public in the US, and he bought in days after it was announced.
Remember: this is the account that flagged $SIVE before +1,200% and $SNDK before +518%. He calls the bottleneck before the crowd. Now he's calling humanoids.
And this isn't a promise stock. $300M+ in real orders, 65,000+ operating hours, customers like Toyota, GXO and Mercado Libre, with Amazon testing. Backers: $NVDA, Amazon, Foxconn, Softbank.
The kicker: it's IPOing at ~$2.5B while Figure sits at ~$49B. He says it feels like ChatGPT the day it was announced.
Early. Cheap. Backed by the biggest names in tech. This is exactly the setup he's gotten rich on before.
We track Serenity's every move and write a full deep dive on each. Link in bio to catch his next position the second he posts it, before the rest of the market wakes up.
You're unable to view this Tweet because it's only available to the Subscribers of the account owner.
Is this the next 100x... or the next SPAC that goes to zero?
Serenity, the account that called $SIVE before +1,200%, just bought $CCXI (Agility Robotics, IPOing as $AGLT in September). The first pure-play humanoid stock in the US.
The bull case writes itself: $300M+ in real orders, Toyota and Amazon as customers, $NVDA and Softbank backing it, IPOing at ~$2.5B while Figure sits at $49B.
The bear case: it's a SPAC, the BOM is still ~$125k, and the deal could slip past November.
So which is it? Humanoid moonshot or hype trap? Drop your take below.
We track every Serenity position and write the full deep dive on each. Link in bio so you're early on the next one, before the comment section figures it out.
You're unable to view this Tweet because it's only available to the Subscribers of the account owner.
GM just replaced 1,000 workers with 50 robots.
And reports say it's now cutting a factory-robotics deal with $NVDA. Serenity's read: this is the robotics inflection point, said out loud.
His take: the tech is already here and the use cases are proven. Robots just aren't at scale yet. The next dominoes are the high-headcount giants, from $AMZN to $GM. They call it "assistive robotics," but it's really about running 24/7 with no salaries and no insurance.
The investable part: when the giants start buying robots at scale, the picks-and-shovels suppliers underneath them re-rate first. He's already long the ones most people can't name, the strain-wave reducers and actuators inside every arm ($688017, $SHA0).
Buy at the close the day he names a stock and hold 30 days: +24% on average. He named $SIVE before +1,200%.
Link in bio for an email the second he names the next one, deep dive alongside it.
$GM cuts 1,000 workers and replaces them with 50 robots (Jun 22nd report)
Reports also show GM is working on a deal with $NVDA on factory robotics.
This is industry validation for robotics, since it likely demonstrates increased opex margins + efficiency from employee automation.
Especially from companies like $AMZN to General Motors with assembly/warehouse lines that have high headcount.
I think the technology is already here and use cases are already proven. Robotics/Humanoids are just pre-scale right now.
The wording from large companies is “assistive robotics”… But in all honesty, it’s here to automate away the human workforce and increase profitability.
I don’t think companies want to be paying insurance, salaries, etc. when robots can do a better job, and 24/7, so very sad to see potential job losses in the future from robotics automation.
Whether people like it or not; seems like the inflection point for robotics is here.
The entire AI buildout is quietly choking on one part: lasers.
Serenity's read from $POET's annual meeting: the top three laser suppliers control 68% of the market and are completely sold out into 2029.
And demand for optical engines is set to go 10x, from about 1 million a year to 1 million a month exiting 2027.
Just some notes on $POET AGM and read through on optical markets:
- "The top three laser suppliers control 68% of the market, and they’re completely sold out for the next two years"
$LITE CEO said into 2028, so POET implicitly confirms laser shortage is going into 2029 now.
- NRE with a new customer, building on POET’s interposer for high-power external light source.
This is high confidence $SIVE as laser supplier given the Sept 29, 2025 PR on ELS, and new customer qualification would be material for revenue if it goes into volume ramp.
- Poet expects Lumilens commercial agreement to scale to over $500 million over the next 5 years.
Lumilens claimed a top-3 hyperscaler was their initial customer (Linkedin OSINT)
- $830M cash on hand on balance sheet (this is more for Poet fundamentals).
- "The entire optical components industry today is facing a severe shortage of critical components."
Reaffirming what we know already regarding optical bottlenecks.
- "Production ramp in the second half of 2026 this year."
Just production timelines H2.
In terms of $POET volumes:
- existing capacity around 1 million optical engines per year
- projected demand exiting 2027 around 1 million optical engines per month
- roughly 10x capacity expansion
I don't own Poet, but if management delivers on these projections it directionally looks very positive.
However something to note is that unlike other suppliers that have named hyperscaler customers to to back up capacity revenues like $AAOI: $POET seems more questionable.
But this does look like a very positive outlook for $POET if they match projections.
While everyone online fights about one AI stock, Serenity just mapped the entire decade.
His 2020-2030 board:
Orbital compute on reusable rockets: $RKLB, $SPCX
AI data-center lasers: $LITE
Directed energy, actual Star Wars beams: $EOS
Humanoids replacing the workforce
Self-driving everywhere: $TSLA, Waymo
Quantum going commercial by 2030
The difference: he isn't theorizing. He's been long these names before the themes went mainstream. $LITE +167% and $RKLB +117% since he first flagged them.
Buy at the close the day he names a stock and hold 30 days: +24% on average.
Link in bio for an email the second he names the next one, deep dive alongside it.
Honestly this decade from 2020-2030 might be the most goated in human history.
- Scaling massive reusable rockets for orbital compute with $RKLB to $SPCX.
- on the cusp of ASI and recursive learning with Anthropic and OpenAI
- Backflipping Boston Dynamics and Unitree humanoids to replace the human workforce
- Star Wars laser beams from $EOS.ASX to AI DC lasers like $LITE.
- Waymo and $TSLA self driving cars everywhere in urban cities
- and we get industry Quantum commercialization end of decade
This is kinda crazy to be an investor in this timeframe. Feels like every movie from the Star Wars Death Star to Skynet is coming to life.
What’s next?
Three months ago, every analyst on your feed said memory stocks were dead money.
An anonymous account on X said buy $MU at $380.
It just hit $1,122.
He doesn't work at a bank. He's not on TV. He posts for free. And the scoreboard isn't close:
$SNDK $565 → $2,155
SK Hynix +216%
Samsung +111%
The whole consensus got smoked by a guy on X. And he's still naming new stocks every week.
His calls average +24% in the 30 days after he posts them.
Link in bio for an email the second the next one drops, deep dive alongside it.
POV: you just found out an anonymous account on X called $SIVE before +1,200%, $MU before +490%, and $NBIS before +200%...
...and he's still dropping new names every week.
His calls average +24% in the 30 days after he posts them.
Definitely worth tracking. Link in bio for email alerts plus a deep dive on every new ticker he names.
Imagine copying one anonymous account's stock calls for a year. Here's what that actually looked like.
Returns since Serenity first named each:
$SIVE +1,203%
$SNDK +518%
$MU +494%
$AXTI +456%
$AAOI +346%
$ARM +214%
$NBIS +200%
No fund. No Bloomberg terminal. A guy posting on X before the crowd showed up.
Buy at the close the day he names a stock and hold 30 days: +24% on average.
Link in bio for an email the second he names the next one, with a deep dive alongside it.
Info only, not investment advice.
The US-China AI race just escalated, and Serenity already told you where the money has to flow.
WSJ: China's Zhipu AI matched Anthropic on cybersecurity benchmarks. His read: the only answer is trillions in capex to defend the AI moat.
That capex doesn't go into thin air. It lands on compute, memory and data centers, the exact names he's been long for months: $NBIS, $MU, $SIVE.
News like this is the catalyst. The positioning came first.
Buy at the close the day he names a stock and hold 30 days, and his calls average +24%. He named $NBIS before +200%.
Link in bio if you want an email the second he names the next one, deep dive alongside it.
WSJ: “China Has Matched Anthropic in Cybersecurity, Resetting AI Race”
China’s Zhipu AI, has matched
Anthropic’s Mythos in certain benchmarks, such as exploiting cybersecurity bugs.
Well, that’s not good.
Instead of blaming the administration which did the correct thing…
Anthropic’s clearly at fault for not setting enough guardrails to prevent distillation.
Even before Fable release there were a lot of rumors that China had backdoors to distill it.
Part of it is the trillions of capex to accelerate the AI moat.
And the second part is not handing that work over on a silver platter for millions in API calls.
Google just quietly admitted it's out of compute.
Serenity caught what that actually means.
$GOOGL restricted Meta's Gemini capacity back in March because it didn't have enough power to go around.
So Meta went and signed massive deals with neoclouds like $NBIS (Nebius) instead.
Read that again: the hyperscalers can't even supply each other. Demand is that far above supply.
That's the entire bull case for the AI data-center buildout, and $NBIS sits right in the middle of it.
Buy at the close the day he names a stock and hold 30 days, and his calls average +24%. He named $NBIS before +200%.
Link in bio if you want an email the second he names the next one, deep dive alongside it.
Guess we finally found why $META signed massive agreements with Neoclouds like $NBIS back in March...
And why Gemini got super nerfed.
$GOOGL reportedly restricted Meta's capacity in March 2026 because of compute restraints.
Google's CEO said from last earnings computing power restrictions prevented Google Cloud from taking on more customer needs and made the department's backlog nearly double the previous quarter.
This is probably positive for the AI DC capex buildout since hyperscalers capacity is way below what is needed, and especially so if they can't rely on one another.
Serenity just found the company that could be half of every humanoid robot, hiding inside a boring auto stock.
$SHA0 (Schaeffler). It already makes the bearings, gearboxes, actuators and power electronics a humanoid needs. His estimate: up to 50% of a robot's entire bill of materials. It's already working with 45 different humanoid players.
Why it's cheap: the market sees a €7.5B auto-parts company and ignores the robotics optionality. Humanoids are ~1% of revenue today. The re-rate comes when that changes.
This is the account averaging ~24% per call (30 days), the one that named $SIVE before +1200%.
Link in bio if you want an email the second he names the next one, deep dive alongside it.
Yup, Schaeffler $SHA0 is the ideal example for automotive players.
~ €7.47B MC auto player.
- but also working with 45 humanoid players (extremely insane number, they do have €5.14b debt, EV is higher)
- covers bearings, gearboxes, sensors/ECUs, actuators, power electronics.
-ests. They can be roughly 50% BOM of a humanoid
- targeting 10% market share of that.
But only projecting a few hundred million revenue by 2030 (extremely sandbagged) if we’re using Elon’s projections for the market.
Then’s theres auto players like Nabtesco (6268) for joints. And then Sanhua in China for $TSLA Optimus.
Many of them are trade at lower valuations from drag from auto.
But humanoids + AI Auto (TSM chairman cited this as a growth vector) should be a good catalyst past 2027 for automotive players.
IMO there just needs to be a ChatGPT/Anthropic breakout downstream leader first for humanoids.
That brings up the entire ecosystem that has robotics exposure.
But as of now it’s probably like 1% of many of their overall revenue figures, and markets are focusing on immediate upside from bottlenecks like memory or MLCC.
I’m sure we’ll see some random Toto-HBM type surprise as different humanoid architecture evolve.
FinXtracker retweeted
We backtested every stock Serenity has named. The result is almost unfair.
Buy at the closing price the day he first names a stock. Hold 30 days. Sell.
Across 35 calls, that simple rule returned +24% on average.
Not his best pick. The average. Every name, mechanically, no skill required.
For context, the S&P returns about that in 3 to 4 years. He does it in a month.
The standouts speak for themselves: $MU +195%, $SNDK +281%, $ARM +214%.
The market still sees $XFAB as an old European specialty foundry.
Serenity sees something else:
NVIDIA / Nokia photonixFAB evaluation
European silicon photonics foundry buildout
InP-on-silicon transceiver platform
TFLN / SOI / SiN integration
SiC + GaN power optionality
EU + U.S. government support
Is X-FAB becoming one of Europe’s open foundry layers for AI infrastructure? If that map converts into volume, this should not be valued like a random automotive-cycle foundry.
Today, $XFAB (1.23B MC) receives €127.4 million from CHIPS ACT for their MEMS Foundry.
"The «Fab4Micro» in Erfurt is a key project for the technological sovereignty of Germany and Europe, explained Thuringia’s head of government."
The production start is planned in 2028 (finanzen).
With Europe's new CHIPS ACT 2 as a concurrent fundamental tailwind.
$JBL is developing a 1.6T LRO pluggable optical transceiver using $SIVE DFB lasers.
Jabil management later described the LRO as having potential for a “relatively dramatic moat”.
The map now looks like this:
$SIVE → $JBL 1.6T pluggables
$SIVE → $GFS silicon photonics reference designs / SCALE
$SIVE → O-Net + Enablence external light sources
$SIVE → $POET optical interposer / ELS modules
$SIVE → Ayar Labs SuperNova light source
Then the OSINT map gets interesting as well:
possible adjacency to Marvell/Celestial, Lightmatter, Lightelligence and other optical-I/O players.
The thesis is not “one design win”.
The thesis is that Sivers might be the laser layer repeatedly designed into the architectures AI infrastructure needs next.
Pluggables.
CPO.
External light sources.
Optical engines.
Foundry reference designs.
That’s a misconception:
$SIVE is the laser supplier for next gen architectures, not just CPO scale up.
Pluggable, scale out CPO, scale up CPO, NPO, etc.
- Sivers and $JBL went god-mode and developed 1.6T optical transceivers with CW lasers.
Effectively designing around the EML bottlenecks, and even made things more power efficient.
From quotes from Jabil management they created a “relatively dramatic moat”.
So for the next gen of 1.6T pluggable transceivers, Sivers seems immediately used.
Markets missed this nuance too: after Jabil’s announcement, other pluggable players reached out, and SIVE is working with them now (prob codevelopment, qualification stage).
There’s not too many players that build optical transceivers that aren’t vertically integrated like Lumentum (think: Innolight/eoptolink, and maybe others).
So these are active developments, just not public material yet and could be a new press release anytime.
- for CPO scale out which happens H2 2026 onward like $POET, $SIVE is a laser supplier for those players.
- scale up CPO is h2 2027, from Ayar and Nvidia’a NVLink CPO ecosystem and that’s the main volume ramp across optical players.
But this is where $SIVE looks like they mog every other player from supply chain mapping.
Since they’ve likely been working + designed in with $MRVL Celestial, Lightmatter, Lightelligence and the others since they were small way back when.
On top of things like this:
For foundries like $GFS, $SIVE is the reference laser.
Not even including massive companies like O-Net building ELS with $SIVE, likely for the Asian hyperscalers supply chains.
This is why $SIVE is by far my favorite laser chokepoint long.
Feels like they’re everywhere in new optical architectures starting beginning of 2027.
Everyone was buying Nvidia.
Serenity bought the photonics bottleneck.
He just turned solid money into a small fortune:
An AI research scientist who saw the constraint before the market did.
He is likely pulling around 60k a month from 1 usd subs alone.
Started with research scientist money; call it a decent base.
The way his investments have gone, he has plausibly 100x’d his wealth.
He has likely made tens of millions this past year.
We track all his moves. Every position, every catalyst, before the market catches on.
Not confirmed. Just what the position implies.
$SIVE $AAOI $XFAB $AXTI
The most interesting thing about Serenity’s posts is the map.
$SIVE
$AAOI
$AXTI
$IQE
$ALRIB
$LITE
$COHR
All different companies.
Same underlying idea:
AI clusters are becoming an optical systems problem.
Copper is bottlenecking AI right now, and photons matter.
When photons matter, lasers, InP, substrates, epiwafers and obscure equipment companies suddenly matter too.
