Just got a text one of the largest companies in the world had a North Korean It operative taken out of their HQ in handcuffs for being a spy...
candidate fraud isn't just a remote problem...
Amazing documentary about North Korea trying to infiltrate US companies by applying for (and getting!) remote jobs by @emmatscott@skesslr@bobmcmillan
Our portfolio company @hiretofu prevents this via AI resume screening and deepfake detection.
https://t.co/pmwRaeQJyY
Curious what a day in the life of a DPRK IT worker looks like?
We researched and documented it all.
Last week I spoke with an investigative journalist on DPRK IT worker fraud.
- It made me realize.. there's an overabundance of reports and articles with the same statistics.
But nobody has gone deep on how the bad actor you end up on a call with gets to that point.
- How are suitable workers are identified?
- How are they procured into the job?
- How are they organized?
- What their day is like?
It's highly deliberate, organized and meticulous.
And for those curious, it's detailed in this weeks blog post. https://t.co/hfDgFijiPJ
Jason Zoltak shares the weekly habit that landed him his lead investor in @hiretofuโs pre-seed round back in 2024. The team has since raised a $5M seed round earlier this year.
CC: @carterkev
Listen to the full convo๐๐ป
Youtube ๐๐ป https://t.co/RE7nS2Gqd5
Spotify ๐๐ป https://t.co/oj1WD4xuyt
Apple Podcasts ๐๐ป https://t.co/VpVU50fPDI
#founderlife #fundraising #venturecapital #investors #seedround #betweentwobrunettes
I think I figured out what makes a really good startup AE..
We closed a tier-1 logo two weeks ago, $10B market cap... it took 2 years.
- Got introduced to VP Talent in August 2024 , one short demo, didn't want to book second call
- Few weeks go by of email exchange
- then another call in October.. crickets after that
- VP leaves to another company
- next, ONE YEAR of follow up emails into the void, 9 emails / cold calls and trying to go through the chimney later... nothing
- then ping the CEO... drives a response from the team
- then crickets again...
All of sudden, something happens in the market and a lightbulb goes off... "I should probably reach out now."
4 days later, new Head of Talent responds
- we book one meeting
- then a second
- and we get the verbal
four weeks later, the contract is signed..
I think an AEs ability to follow up until they d** but also have their head on a swivel to listen for clues in their everyday that signal to them to reach back out...
are two of the most important things you can have as a startup AE.
thats not the only deal we closed like it.
and I think its what separates the good from the great.
We sat down with @JasonZoltak, Co-Founder of @hiretofu, to talk through how customer feedback led him straight into a $96 billion fraud problem.
In this conversation we get into the pivot that changed everything, the North Korean IT worker scams targeting hiring teams, and what it took to build a fraud platform.
Full episode with Jason out now! Links in the comments๐๐ป
#tofu #fraudprevention #hrtech #recruitingtech #startupfounders #cybersecurity #jasonzoltak #betweentwobrunettes
We put the name's of the people who tried espionage-ing tofu in this week's Brian episode :)
Candidate fraud is a highly coordinated operation.. but even then, fraudsters mess up and expose themselves.
Even those who have tried to espionage Tofu are sloppy and end up revealing themselves.
The tells are either easily visible or hidden beneath the surface.
But they're there, if you know where to look.
It's a core part of our fraud detection product and why our check accuracy leads the market.
This week's episode of Brian takes us through exactly that...
Enjoy
Not all deepfake software does the job you need it to.
The best deepfake detection software scores 98% in a lab.
Research shows why those models with amazing results fail on a normal every day video call.
First, take a compressed video -- the ordinary squeezing every video conf platform does to save bandwidth. Accuracy drops from 98% to 61%.
Second, show it a fake built by a tool it had never seen. New deepfake technology ships every month. It scored 52%. A coin flip.
Third, add a layer of static so faint no human eye can see it. The detector went from 98% to 0.2%.
A change you physically can't perceive takes the software from near-perfect to terrible.
Because it's not purpose built for interviews.
The interview setting is imperfect.
A model built to display accuracy fails in the room that matters.
Research if you're interested - https://t.co/9JE1idr6Nq
The part of North Korean IT worker fraud that still blows my mind:
There are managers that oversee everything.
They are not rogue freelancers independently applying to remote jobs.
- Itโs an operation.
- Managers run teams.
- Workers have revenue quotas.
- Top performers get rewarded and promoted.
There are delegation leaders and group leaders managing the workers below them.
One DOJ case outlined an operation with at least 130 IT workers that generated $88M (link in the comments).
Managers were responsible for ensuring workers generated at least $10,000 per month.
- They have an org chart.
The more we learn about applicant fraud, the clearer it becomes: companies aren't fighting individual bad actors.
- They're fighting businesses built to get people hired.
- Meet the manager.
tofu signed its largest six-figure contract ever yesterday.
it's crazy to say we now protect the recruiting pipelines of companies that likely every other tech company uses in their stacks.
and last night we took an hour to celebrate
We are also launching some major feature additions and have added incredibly talented people to our team help our existing and new customers fight fraud:
- 16 new signal categories which include hundreds of sub-signals to catch fraud rings and other sanctioned groups
- connectivity between applicant fraud and deepfake detection
- ML eng talent from the fraud teams of block and cash
fraud is still at its early onset.
we're working harder than ever to become the standard protection layer in recruiting and hiring.
reach out if we can help... or even just chat to share what we're seeing and how we can help :)
Is giving a sh*t a competitive advantage? I'm starting to believe it can be.
We won a big RFP for our deepfake product last week. We beat a number of the seemingly top companies on the market.
And among the reasons we won was.. in the customers words.. our commitment to going deep with their team.
It made me think back to the other opportunities we've earned because we really truly care.
For instance tofu's first fundraise: When we raised our first round, weeks later I was with our investor Chris Howard at the SF Giants game. I asked him why he invested in us.
- he said "nobody I've ever invested in responded to my emails faster than you."
Other deals we've closed I've stayed on those for literally two years.
Two years of consistent follow ups, sharing updates... which ultimately led to doing business together.
Because @jonathanperez_0 and I truly care.. and believe in what we're doing.
All else equal, what really moves the needle?
People buy from people.
I think subconsciously we can sense this.
I believe it creates an edge.
Or at least it feels that way.
If you're using your ATS built in candidate fraud detection, read this...
3 months ago, a prospect chose their ATS's built-in candidate fraud feature over Tofu.
- We closed out the deal.
Yesterday, we met with them again because they hired a North Korean candidate. Their ATS greenlit him.
On the call, our AE pulled up our system. We'd already flagged that candidate. Over 100 times.
We hear a version of this story every week.
Not all fraud detection is built the same. Many assume it is.
Outsourced, non-recruiting fraud specific solutions decay over time.
That decay is happening now, inside tools people trust.
Brian, episode 2 "The Operation" is out now...
Many people are surprised to learn candidate fraud is an end to end operation.
So here's Brian to explain more...
Enjoy. @hiretofu