Croma Platform Analysis
An assignment given in one of the interviews I sat in.
The product is stuck between the clinic, the code, and the devices.
I work with founders who need someone who has sat in all three rooms. The clinic, the software, and the hardware. Not a generalist PM. Not an agency slide deck.
01 — Work
Outcomes from roles, not a brochure.
Numbers below come from products I helped ship at Cavista, Numen Health, Navya, and as an independent consultant at WiseRx.
Cavista
EMR interoperability
42% exchange · 60% less manual · 15+ systems
Problem
12+ healthcare systems could not exchange patient data reliably. Care teams reconciled by hand.
What I did
Mapped flows, put FHIR/HL7 in as the exchange layer, sequenced the noisiest integrations first. Product role.
Cavista
Stability monitoring
47% less downtime · 4.2h → 1.8h MTTR
Problem
Incidents arrived from users first. Root-cause across the portfolio was slow and reactive.
What I did
Prometheus and Kibana with anomaly signals and playbooks so the team could act before the ticket pile. Product role.
Numen Health
AI-powered apps, adoption not theater
40% adoption · 20% satisfaction · $2.8M ARR
Problem
Need to launch AI mobile apps for 20K patients and 500+ providers without knowing which features would actually get used.
What I did
Roadmap from instrumentation, not opinions. Fast cycles. Product role.
Navya
HIPAA messaging
4h → 15m response · audit pass
Problem
Doctors and patients had no fast, compliant channel. Response times stretched to hours.
What I did
End-to-end encrypted in-app messaging, built to pass a HIPAA review. 2,500+ doctors, 50K+ patients. Product role.
WiseRx · now
Hardware software
FDA QMS · process building · firmware-to-cloud
Problem
A medication-management device had to qualify as Medical Device Software — not an app glued onto hardware.
What I do
Consulting on SDLC, QMS, process building, the firmware-to-cloud path, and two-way device ↔ patient communication. Current consulting.
Also: onboarding work with ML / multivariate testing (30% conversion, 45% less drop-off) on a health platform — scoped product work, same stack of habits.
Selected surfaces
Consolidated Mockups
Various assignments clubbed from the interviews I sat in.
Optimizing existing Field Collection Process
An assignment given in one of the interviews I sat in.
Example — Live stack
A practice running on a thin full stack.
Thin frontend and backend, no standing engineering team, still a product someone can run. Built and kept stable with agentic AI.
How it is built
Product
The thin site or portal the client actually uses.
Build
Agentic AI, directed by me.
Data
Business records live in DynamoDB.
Hosting
The frontend runs on AWS Amplify.
I review changes before they go live. The client runs day-to-day ops. No standing engineering team.
How AI is used
Used when the work calls for them. Not a checklist on every product. Generative AI, RAG, or agents, then a check, then a person, before it ships or acts.
Generative AI
Models draft, summarize, or produce product content.
RAG
Answers stay grounded in the client’s own documents.
Agents
A model uses tools and finishes a task, not only a chat reply.
Evaluation
Outputs are checked before they reach a user or an ops desk.
Human in the loop
I or the client approve before it ships or acts.
Hey Upasna
Clinical psychology practice on the web. Minimal frontend and backend. Live and kept running with agentic AI.
Cause & Case admin
The ops team runs the non-profit from this portal. No engineering staff. Built, managed, and kept stable with agentic AI.
02 — Offer
You hire a person, or you hire a lead with a small pod.
Same method: understand, ship, measure. Different size of problem. Price is scoped after a call — not a rate card on this page.
Project
Me, on a defined problem.
Weeks. Inside your team. One owner of the brief.
- Roadmap, AI feasibility, IoT / device audit, a v1 slice, a pilot.
- You talk to me. No account manager.
- You leave with a document, a shipped slice, or a working prototype.
Lead
Me, plus people I bring.
A quarter or more. One accountable lead.
- Initiative too big for one specialist — strategy and execution under one cadence.
- I coordinate or embed with your team when the work needs them.
- Weekly reporting. Blockers are mine until they are not.
A fit if
- You are health-tech. Software, clinical workflow, or a device that talks to software.
- You need judgment in the product, not extra Jira traffic.
- You will share constraints: budget, compliance, hardware, timeline.
Not a fit if
- You want a generalist growth or brand studio.
- You need a staffed 20-person shop pretending to be one consultant.
- The work is outside health-tech.
03 — Practice
Operator, then founder, then hired gun.
I am Adarsh Kumar Tiwari (Adarsh Tiwari), founder of Adarsh & Consultants. I have worked the clinical side, the software side, and now the device side, which is why I am slow to romanticize features. Product at Navya, Numen Health, Cavista. Current consulting on a smart medication device at WiseRx.
I co-founded PillSwift in 2024, an online retail pharmacy and wholesale delivery platform for Tier 2 and Tier 3 cities in India. Early signal was weak, so I halted working on it.
I co-founded Cause & Case in 2026, a Section 8 non-profit for free pro bono legal aid in Delhi NCR and Madhya Pradesh. We take the cause first, then the case, at no cost.
McKinsey Forward, MIT Sloan product design, IBM Scrum, HIPAA work at Navya and Cavista. Useful, not the pitch. The pitch is pattern-matching in this domain.
04 — Words
From people who managed or sat next to the work.
LinkedIn recommendations, excerpted.
Vardaan Vashisht Entrepreneur & investor
I had a delightful experience working with Adarsh. He is a quick learner and took initiative on his own to go beyond the responsibilities assigned to him. Very open and easy to communicate with.
Dr. Judith John Pharm D, RPh
Adarsh is an enthusiastic person with so many inbuilt talents. I had the opportunity to work with him on ISPOR student chapter activities — it was wonderful working with such a motivated person. He deserves my highest recommendation.
Ana Beatriz Pizzaro Evidence Synthesis Groups Officer, Health Equity
We worked together on the Chapter Development Committee for a year and it was a great experience. Adarsh has very good communication and decision-making skills — I would work with him again without a doubt.
Dr. Niti Mathur MBBS, DGO, SMP (IIM) · managed Adarsh at Navya
Adarsh has worked with Navya Care for over a year, and I was personally happy to note his commitment, enthusiasm and engagement with the organization. He executed well in support as well as clinical operations.
Diwesh Jaiswal Director of Engineering, AI & ML, Navya
Adarsh is a very enthusiastic and skilled individual. He has taken very little time and scaled up in his role as a product analyst. I can see him growing into his current role very rapidly and being an asset to the organization.
Pooja Akhileshwar Senior Manager, Patient Services · managed Adarsh at Navya
He was found to be a very empathetic analyst with fair clinical knowledge. As a fresher in the field of Project Management, Adarsh exhibited an impressive eagerness to learn and a natural ability to grasp new concepts quickly.
Armugam M Lead QA Engineer, Navya
Adarsh is a quick learner, dedicated, and hardworking. He joined Navya as a Clinical Analyst and grew into a Product Analyst role. I observed him to be dynamic and enthusiastic about every task given — a true multi-tasker and team player.
Pranay Mahajan Product, Navya
Adarsh consistently showcased exceptional analytical skills, ensuring a deep understanding of client needs. His dedication to refining these requirements played a pivotal role in our team's accomplishments.
Sushruta Marla Junior Product Manager, Navya
We worked together at Navya Care as clinical analyst and later as product analyst. He did great job at whatever work was assigned to him. He is curious to explore many things, which makes him different from others. He is a team player.
Barsha Pattnaik CSPO, health-tech / AI product
His strong product management skills and data driven approach were evident. He brings clarity and structure to discussions, and I'm confident he'll add value to any team that he joins.
1 / 2
05 — Straight answers
What founders ask before the first call.
Only health-tech?
Yes. FHIR/HL7, HIPAA, clinical workflow, device software — that is the pattern library. Everything else I decline.
Pre-seed?
Yes, as a scoped project: validate direction before you fund a bigger build.
NDA?
Yours or mine, before any detailed discussion.
Cost?
Scoped after a call. Project and lead land in different ranges. A number within a day of scoping — not a guess on the website.
If it is not working?
Either side can stop. Projects use short checkpoints. Lead has a trial window before a longer commit.
Who owns the work?
You do. In the agreement before we start.
06 — Ask
Send the problem. I reply within 48–72 hours.
Pick a time if you already know you want a call. Form if you want it in writing first.