@twitharshili
iAccount based inIndia
About this account
- Account based in
- India
- Connected via
- India App Store
Account-level information from X, not a live location or the device used for a specific post.
Software Engineer @Deel 💸 | Ex-Tech Lead @getPostman 🚀 Taming AI Agents, global payroll code, & breaking APIs | Hot takes, tech politics, & system edge cases
Joined April 2013
- Tweets126
- Following333
- Followers172
- Likes69
The hardest part of AI in regulated work isn't building it. It's earning trust.
At @deel , AI now scores workers' comp risk for every new hire. Low-risk roles clear automatically. Experts get the hard cases.
We check it against our underwriters daily: 86% agreement so far.
Proud to be recognized as a Deel Values Champion for "Exceed Expectations" in Q3'26 🏆
Key highlight of this quarter was I got a chance to build an AKAI (akai.run) AI agent that triages PEO payroll sync issues and saves 900+ engineering hours a year. I wrote more about the agent here
deel.com/blog/900-engineerin…
Great thanks to @deel for the nomination and all the support!
🤖 Made with AI
If anyone interested, let me know 😊
We're hiring Deel's first Chief Scientist.
The labs are making models smarter. We're building what comes after: AI that does high-stakes work reliably, at global scale. Payroll that can't be wrong. Agents that act, not just answer.
You'll set the research agenda, build the benchmarks, and stay hands-on, with data on how 150+ countries work and a research-to-production loop measured in weeks.
Reports to the me directly! Know someone? I'd love to chat.
Let’s go 🚀🚀🚀
From the Bay to the World 🌎
Deel is proud to be the Official HR + Payroll Partner of @warriors, @Valkyries, @ChaseCenter and @gleagueWarriors.
AI isn't replacing global HR teams—it’s elevating what entry-level ops looks like. 📈
If you’re only using AI to draft emails, you’re missing 90% of its leverage. Here’s how high-growth teams incorporate AI + @Deel into global ops:
1️⃣ Assistive (Drafting & Research):
Using Deel AI to instantly check localized labor laws, statutory PTO, or benchmark salary ranges in 150+ countries. (Saves ~20% time)
2️⃣ Augmentative (Structured Analysis):
Running multi-currency FX reconciliations, flagging tax compliance risks, and auditing cross-border payroll runs automatically. (Saves ~50% time)
3️⃣ Autonomous (Agentic Execution):
Deploying AI agents—like Akai by Deel—to handle end-to-end back-office workflows, background compliance, and device shipping with human-in-the-loop signoff. (Saves 80%+ time)
The bottleneck is rarely the technology—it’s how clearly you define the workflow before delegating it.
Which level is your ops team operating at today? 👇
If your hiring strategy is limited by borders, you're missing out on 99% of the world's talent pool.
@deel makes managing a global workforce feel as simple as managing a local team:
• Onboard international hires compliantly in minutes
• Run global payroll across dozens of currencies with total transparency
• Manage EOR, contractors, and internal HR from one single source of truth
Borderless work is no longer the future—it’s the default operating model.
Are you hiring locally or building a distributed team? Let’s connect in the comments 👇
#FutureOfWork #GlobalTalent #HR #Deel #WorkforceManagement
Most AI tools promise automation, but building them takes months of engineering tickets.
AKAI (akai.run/) by @Deel changes that. Here’s how ops teams are using it to automate back-office workflows in 3 steps:
1️⃣ Record Once: Show Akai your workflow once or describe it in plain language. It maps the steps and builds connectors on any web system—no API needed.
2️⃣ Deploy AI Agents: Set up autonomous agents for complex tasks like FX reconciliation, multi-country payroll preparation, and tax portal filings.
3️⃣ Human-in-the-Loop: Every action is fully logged, audited, and human-reviewed, getting smarter with every single run.
Saved 91,000+ hours/month at Deel before going live.
Ready to put manual operations on autopilot? 🚀
Software ate the world.
AI agents are eating software.
SaaS sells seats → Agents sell outcomes
SaaS automates screens → Agents automate decisions
SaaS requires training → Agents observe and learn
We built AKAI (akai.run/) at @deel for exactly this.
91,000+ hours saved/month. Not a better tool — no tool at all.
Software is a tool. Agents are a workforce.
#AI #Deel #Akai #FutureOfWork
Regulation does not build the future. Builders do.
Super intelligence is not a threat to manage—it is the cornerstone of 21st-century sovereignty, security, and human freedom.
Nations that embrace it will lead. Those who hide behind global bureaucracy will stagnation.
We are at a defining crossroad. The future belongs to those brave enough to shape it.
"A prosperous future will not be secured by a global regulator. It will be secured by sovereign nations that adopt super intelligence, responsible companies that build it, and free people who refuse to be ruled by fear."
youtube.com/watch?v=vZ6Bgd94…
True security in the age of AI will not come from global bureaucracy or defensive over-regulation. It will come from technological leadership, national sovereignty, and courage.
Today at the UN Security Council, It seems like the case for a future defined by progress, not permissions.
Payroll failures used to take me 45 minutes each to debug.
I built an Akai agent that does it in under 10 seconds.
Here's what it does:
→ scans and groups failures automatically
→ retries instantly (resolves 55% on the spot)
→ fans out queries across 4 systems in parallel
→ classifies root cause + generates exact fix commands
→ drops a Slack message with copy-paste fixes
Result:
- 900+ engineering hours saved per year
- 4-6 cases handled → 20+ simultaneously
- I went from reading queues to reading a Slack message
I wrote the full breakdown here: deel.com/blog/900-engineerin…
Harshil Jain retweeted
EXCITED TO LAUNCH: Akai (akai.run)
Deel added >$140M ARR in 90 days without increasing headcount by automating~600 Full Time Employees' equivalent in work with Akai.
Akai was an internal tool to automate our painfully repetitive operations in Finance, HR, Accounts Payable, and Compliance, etc. We never intended to make this a product.
But we watched revenue per employee grow from $130K to $215K
We built >8k agents that do the work of ~600 employees
It had such a dramatic impact on our business that today we are launching it for everyone.
How it works: Say you're automating payment reconciliation:
1. Record your screen while manually matching a messy transaction and Akai will capture your screen, voice, server requests
2. Akai will see that you pulled unformatted wire transfer info from an archaic bank portal, put it in some excel sheet, checked NetSuite invoices, payment history, and put a ticket on Zendesk
3. Akai reads between the lines and build a workflow + steps + conditional guardrails. It learns tacit edge cases, like resolving malformed invoice references without you writing a single regex
4. Simply connect NetSuite, your ledger, Zendesk, PSPs, and even legacy bank portals with zero API access
5. Run the workflow and tell it what to adjust in plain English: "strip slashes on wire memos and auto-apply partial payments." It adapts instantly
6. Once it works for you, add 100s of colleagues. Your entire payment ops team forks and extends the workflow for new PSPs, secondary ledgers, or regional settlement rules
7. We automated 85% of our payment reconciliation end to end, eliminating 500+ hours of soul-crushing manual grunt work every single week.
Claude Code/Codex can't do this in multiplayer mode. Every person rebuilds the same skill from scratch in their own way.
Deel built Akai to:
1. Understand backend operations edge cases (it had to work for our 7000 person team first)
2. Collaborative across 1000s of employees
3. Self-Learning from millions of runs
4. Optimises cost and gets cheaper every run
We're so confident that we're announcing an Automation Guarantee:
If our engineers can't automate a thousand of hours of work in your first 30 days, you get a full refund.
Book a demo: akai.run if you're an exec at a company with hundreds of employees
The ROI on AI-assisted security is wild:
• 86.2% on CyberGym benchmark
• 47.2% on CWE-Bench patching
• 70%+ vulnerability discovery rate
Finding bugs is great, but automated patching at Flash-level pricing is where the real leverage is 🛡️🚀
We’re also introducing Gemini 3.8 Flash Cyber, our most capable cybersecurity model. It shows frontier-level performance in discovering vulnerabilities and patching them at scale, with Flash-level speed & pricing.
That includes achieving 86.2% on the important CyberGym industry benchmark, plus 47.2% on CWE-Bench for patching. We saw a 70%+ success rate in discovering vulnerabilities across 20 programming languages on our internal benchmark.
The highest ROI of AI isn’t generating code. It’s eliminating noise that stops engineers from building.
At @deel , a failed payroll sync cost 45 min of detective work across 4 tools. 55% were just transient glitches — ghosts that disappeared on retry.
We built a 4-agent pipeline on Akai. Now they self-resolve in seconds.
900+ eng hours reclaimed/year.
Read more about it here 👇
deel.com/blog/900-engineerin…
I read both the bullish and bearish takes on AI.
The positive news gets attention. The negative news gets dismissed.
But you need both.
If OpenAI really runs out of cash in 2027, that wouldn’t just be an OpenAI problem — it could expose how much of the AI boom is built on spending ahead of sustainable economics.
Don’t be blindly bullish.
Don’t be blindly bearish.
Read both sides. Then form your own conviction. 🧠
Applications are open, go check it out quickly 🔥
Applications are officially OPEN 🍦@snoopdogg and @realdrbombay are hiring an International Ice Cream Taster, and you can apply from anywhere in the world through Deel. 🌎
I don’t think people realize what Anthropic is actually claiming.
A $30 trillion addressable market is roughly the size of the entire U.S. economy.
That’s not just a business opportunity.
That’s a claim on a massive share of the global economy.
JUST IN: Anthropic is reportedly expected to tell investors its total addressable market exceeds $30,000,000,000,000.00.
Very very useful feature it will be 🔐🔐
ChatGPT Work can now use its computer and browser to sign in to websites on web and mobile, without ChatGPT ever seeing your username or password.
That means you can ask it to:
• Set up utilities for a new apartment
• Book a DMV or passport appointment
• Check reimbursement costs through your insurance
• Find and book an in-network doctor around your availability
• Check when your car registration expires and prepare the renewal paperwork
• Compare your rental insurance policy with an issue you’re emailing your landlord about
• Find and save apartment listings that match your criteria
• Restock something just by uploading a photo
• Schedule a package pickup for a return
• Cancel tickets for a rescheduled trip
• Book a vet appointment
• Submit reimbursement paperwork for medical treatments
• Check resale sites for new drops and save things you might like
• Find candidates with specific experience and draft outreach
• Take invoices from your email and submit them to your accounting software
• Draft replies to rental property inquiries
• Fill out permit applications for your small business
• Add action items to a vendor portal based on a recent client call
• Analyze the latest ad campaign for your small business