Transforming critical institutions with AI.

Joined April 2026
Photos and videos
Curious what applied AI engineers do at Percepta? “Day to day, my role might be talking to doctors, talking to nurses, seeing what they do, mapping out a patient as they go in the hospital. Or it might be just sitting down and coding,” says Anushka. If you're interested in solving real world problems, we're hiring and we want to hear from you.
RajvMehta and Anushka Gupta showed up at a hospital and didn’t write a single line of code for days. They were following nurses through their shifts. Sitting with operations staff. Identifying bottlenecks and constraints through observation — things that would have otherwise never made it into a brief. That’s what Raj and Anushka of @percepta do as Applied AI Engineers. They embed inside organizations — hospitals, governments, financial services, manufacturers — understand their operations, and build AI tools that actually deliver transformation. The work is more fieldwork than ML. More listening than coding. Scoping, Raj told our Builders, is the hardest part: "Nowadays, building software is becoming easier and easier. The real challenge now is…[to] really consider: what is the problem, and what should I make?” One example? Nurse scheduling that took hours to set manually each week was replaced with a streamlined tool so the nurse manager can focus on the hard calls instead. The routine work, handled. The human judgment, preserved. “This job will be on the rise,” Anushka said. “And you guys are getting in on it very early on.” As Forward Deployed Engineers, Pursuit Builders are already doing this work. Many came in with years of frontline experience in healthcare, education, construction, logistics. That experience isn’t a detour. It’s a head start. Full recap at the link: pursuit.org/blog/applied-ai-… Want to inspire the next gen AI-native talent? Be the next fireside chat speaker. Email: Devika@pursuit.org #pursuit #appliedai #forwarddeployedengineer #ai #futureofwork
2
2
5
803
Percepta retweeted
Aka join @percepta!
Once again @sytaylor tells people where the puck is going. Our B2B customers dating back to Q4 confirm that companies want “neutral FDEs” and infra.
1
4
21
5,007
Reupping this seeking architects in new york. Come build the future w us. Healthy referral bonus to anyone who we hire as a staff platform eng :)
2
393
🗞️ Thank you to @washingtonpost and @RepDerekKilmer for highlighting our work with the State of Maryland and @AnthropicAI to help people get access to benefits faster and speed up the permitting process to build more housing.
3
1
7
728
If you're energized and excited by this kind of work, we'd love to hear from you. jobs.ashbyhq.com/percepta
1
1
215
Our founding research scientist, @ChristosTzamos, outlines how we built a computer INSIDE a transformer that can run programs for millions of steps in seconds, solving even the hardest Sudoku with 100% accuracy.
"Our collective, sector wide ambition is being constrained by our own behavior.” @htaneja's inaugural quarterly review is a reflection on building the AI era with our souls intact. It explores @percepta's LLM-as-computer breakthrough, the Customer Value Fund reaching investment grade, our $5 billion commitment to India over the next 5 years, and more.
1
2
10
879
Everyone's racing to build the next frontier model. We think the bigger opportunity is making sure the ones we already have actually reach the industries that need them most. 🧵Our CEO @Hirsh on what Percepta is building, why it's important, and why you should join:
7
5
13
1,322
You should join us. We're in NYC, Boston, SF, and London. jobs.ashbyhq.com/percepta
1
4
245
Replying to @hirsh
Same playbook, different industries: - Improving access to gov benefits in Maryland with @AnthropicAI - Disruption-handling for global supply chains - Voice agents for at-risk patients (w/ Cityblock Health) - Transforming functions for finserv @JHIAdvisors
1
6
212
Replying to @hirsh
Then agents take over the legwork: texting, calling, booking, following up. Every interaction teaches the system which channels and cadences work for each patient. Bottom line? Clinicians get their time back for the real, human work of care.
4
155
Replying to @hirsh
Take healthcare. For hundreds of thousands of Medicare & Medicaid patients with chronic conditions, our system flags what each person needs most: an overdue cancer screening, a ride to the hospital, a prescription about to run out.
1
5
248
Replying to @hirsh
This kind of AI is historically gatekept to corporations with in-house research teams. We're changing that. We build models that surface the decisions driving your business, map them to where work gets done, and run agents that handle it.
4
151
Replying to @hirsh
Every business runs on decisions: what to do, when, how, and why. Figuring that out is hard, and not everything should be automated. But automating the right parts lets companies do more with their resources, and focus on what they do best.
4
140