Dhriti Kalpa Saha
Senior Full-Stack & AI Engineer — ten years of AI-enabled platforms and high-performance SaaS
Senior full-stack and AI engineer with 10+ years building and scaling AI-enabled enterprise software and high-performance SaaS products across healthcare, fintech, education and defense.
I am a senior full-stack and AI engineer based in Agartala, India, with more than ten years spent building and scaling AI-enabled enterprise software and high-performance SaaS products. Most of that work has been in healthcare, fintech, education and defense.
My recent work is AI-first. At Appato I architected Takshala, an analytics platform that gives schools natural-language access to their own institutional data, and built the core AI engine that integrates structured and semi-structured datasets into decision-support workflows. I also led work on financial analytics and portfolio dashboards for hedge fund and prop-trading use cases, and on an orchestration protocol built with ERC-8004 Trustless Agents that lets a fund delegate cross-chain financial operations to independent AI agents.
I have also served as a forward deployed engineer at DRDO-DMDE in Hyderabad, building an air-gapped, knowledge-graph-based RAG platform with end-to-end document ingestion, OCR, image parsing, chunking, embedding generation and hybrid retrieval. It scored 0.86 on the RAGAS scale.
I started out in 2015 building web platforms — a blogging product for photographers at Crive, a registration and billing portal at Vinicom Broadband — and moved through a healthcare CMS at Klinify and a connected-device product at Kesh AI before returning to AI-first platform work. I studied computer science at the National Institute of Technology, Agartala.
Based in Agartala, India.
Selected work
Takshala (2023 — Present)
Senior Staff Software Engineer, Appato
An AI-first analytics platform that gives schools natural-language access to their own institutional data.
Takshala is an AI-first analytics platform for schools. It gives staff natural-language access to institutional data and generates insight across academic, operational and administrative workflows, rather than requiring anyone to know their way around a BI tool first.
I architected the platform and built the core AI analytics engine, which integrates structured and semi-structured datasets into decision-support workflows. A good deal of what used to be assembled by hand now comes out of the platform instead — the work involved in manual reporting and analysis dropped by 60%.
Outcome: Manual reporting effort down 60%
Technologies: AI analytics, NLP, Decision support, Education, Platform architecture
Air-Gapped Knowledge Graph RAG (2026)
Forward Deployed Engineer, DRDO-DMDE
A retrieval platform built to run with no network access at all — ingestion through to hybrid retrieval.
The brief was a retrieval platform that could operate air-gapped, where nothing may call out to a hosted model or a managed vector service. I built it end to end: document ingestion, OCR, image parsing, chunking, embedding generation and hybrid retrieval, backed by a knowledge graph.
It cut the time taken to produce a report by half. The number I care about more is 0.86 on the RAGAS scale, which for a system like this is the one that counts — it says the retrieved context was accurate and that the answers stayed faithful to it.
Outcome: 0.86 on the RAGAS scale
Technologies: RAG, Knowledge Graphs, Hybrid retrieval, OCR, Embeddings, Air-gapped, Defense
Trustless Agent Orchestration (2023 — Present)
Senior Staff Software Engineer, Appato
An on-chain protocol letting a fund delegate cross-chain financial operations to independent AI agents.
The problem was trust. A fund wanted to delegate complex, cross-chain financial operations to specialised, independent AI agents — but doing so normally means trusting either a human manager or a third-party black-box model, and neither was acceptable.
I built an orchestration protocol using ERC-8004 Trustless Agents, implementing on-chain Identity (ERC-721), Reputation and TEE Validation registries so that machine-to-machine commerce could be verified rather than assumed. It supports more than 10,000 autonomous agent interactions.
Outcome: 10,000+ autonomous agent interactions
Technologies: ERC-8004, ERC-721, Agentic workflows, TEE, Blockchain, Fintech, Solidity
Hedge Fund Analytics Platform (2023 — Present)
Senior Staff Software Engineer, Appato
Financial analytics and portfolio dashboards for hedge fund and prop-trading desks.
I led development of financial analytics platforms and portfolio dashboards built for hedge fund and prop-trading use cases.
The platforms include automated trading and analysis systems supporting strategies that manage over $50M in daily assets.
Outcome: Strategies managing $50M+ in daily assets
Technologies: Fintech, Analytics, Portfolio dashboards, Automation, Trading
Cancer Management & Symptom Tracking (2023 — Present)
Senior Staff Software Engineer, Appato
A patient-facing platform for symptom tracking, monitoring and care coordination.
A cancer management and symptom tracking platform, adopted by more than 10,000 users.
It improved patient monitoring, workflow visibility and care coordination through data-rich application flows.
Outcome: Adopted by 10,000+ users
Technologies: Healthcare, Patient monitoring, Care coordination, Clinical workflows
Hair Health Companion (2022 — 2023)
Staff Software Engineer, Kesh AI
The companion app for a hair-scanning device, taken from emerging concept to production.
The device did the scanning; the app was where the value actually landed. I spearheaded development of the companion application, delivering workflows for tracking hair health, care regimens, symptoms and treatment effectiveness.
I integrated data-backed and AI-assisted insights into the product experience, which helped improve regimen adherence by 40% through personalised journeys and clearer progress visibility. I also worked on the device SDK and the connected application workflows, which went on to support 10,000 concurrent device-user connections.
Much of the job was translating an emerging health-tech concept into something scalable and production-ready — hands-on across backend design, APIs, application logic and technical delivery.
Outcome: Regimen adherence up 40%
Technologies: Healthcare, Device SDK, Personalisation, Connected devices, Consumer app
Skills
AI & ML
- LLM applications
- RAG systems
- Agentic workflows
- Knowledge Graphs
- Prompt Engineering
- Context Orchestration
- MLOps
- Computer Vision
The core of my recent work — the DRDO-DMDE retrieval platform scored 0.86 on RAGAS.
Software
- React
- Next.js
- TypeScript
- JavaScript
- Node.js
- Golang
- Python
- Rust
- Solidity
- GraphQL
- REST
- gRPC
Infrastructure
- Docker
- Kubernetes
- AWS
- GCP
- Jenkins
- CircleCI
- SQS
- SNS
- Kafka
- Sentry
- Datadog
- Observability
- GitHub Actions
Data
- PostgreSQL
- MongoDB
- SQLite
- Redis
- Neo4j
- Airflow
- Kafka
- RabbitMQ
Neo4j underpins the knowledge-graph retrieval work.
Experience
Senior Staff Software Engineer, Appato
2023 — Present
- Lead multidisciplinary engineering teams across AI, fintech, healthcare and enterprise platforms, maintaining a 95% on-time delivery rate across client and product engagements.
- Architected Takshala, an AI-first analytics platform for schools, enabling NLP-driven access to institutional data and insight generation across academic, operational and administrative workflows. Built the core AI analytics engine, integrating structured and semi-structured datasets into decision-support workflows that reduced manual reporting and analysis effort by 60%.
- Led development of financial analytics platforms and portfolio dashboards for hedge fund and prop-trading use cases, including automated trading and analysis systems supporting strategies managing over $50M in daily assets.
- Built an orchestration protocol using ERC-8004 Trustless Agents, implementing on-chain Identity (ERC-721), Reputation and TEE Validation registries to enable verifiable machine-to-machine commerce across 10,000+ autonomous agent interactions — letting the fund delegate complex, cross-chain financial operations to specialised, independent AI agents without having to trust human managers or third-party black-box models.
- Delivered a cancer management and symptom tracking platform adopted by 10,000+ users, improving patient monitoring, workflow visibility and care coordination through data-rich application flows.
- Designed workflow orchestration and reporting automation systems for healthcare clients including Enna Health and preventive health startups, eliminating repetitive manual processes and saving 100+ operational hours per month.
Forward Deployed Engineer, DRDO-DMDE
Apr 2026 — Jun 2026 — Hyderabad
- Built an air-gapped knowledge graph-based RAG platform with end-to-end document ingestion, OCR, image parsing, chunking, embedding generation and hybrid retrieval pipeline, reducing report generation workflow time by 50%.
- Achieved 0.86 on the RAGAS scale, proving high context retrieval accuracy and faithful, relevant answer generation.
Staff Software Engineer, Kesh AI
2022 — 2023 — Singapore
- Spearheaded development of a companion application for a hair scanning device, delivering workflows for tracking hair health, care regimens, symptoms and treatment effectiveness. Integrated data-backed and AI-assisted insights into the product experience, helping improve regimen adherence by 40% through personalised user journeys and clearer progress visibility.
- Collaborated on development of the device SDK and connected application workflows, enabling 10,000 successful concurrent device-user connections.
- Worked hands-on across backend design, APIs, application logic and technical delivery, translating an emerging health-tech concept into a scalable, production-ready platform.
Senior Frontend Engineer, Klinify
2018 — 2022 — Singapore
- Led the frontend team on a healthcare CMS deployed across thousands of hospitals and clinics in Southeast Asia.
- Modernised a legacy codebase, improving maintainability, resilience and long-term development velocity by 30%.
- Optimised core web vitals and the observability stack, contributing to a 35% reduction in software-related downtime across deployments.
- Implemented major product capabilities including collaborative features, offline-ready workflows for low-connectivity environments, and template-based reporting and documentation systems.
Full Stack Developer, Crive
2015 — 2017
- Developed a customisable blogging platform for photographers using the MERN stack, focusing on performance, extensibility and user experience.
- Built 30+ complete themes and plugins to enable rapid customisation and monetisation for creative users.
- Ported the web application to a native Android app using Kotlin, expanding platform reach to mobile users.
Platform Engineer, Vinicom Broadband
2015
- Developed a client registration and payment portal to streamline onboarding and billing processes for company customers.
- Maintained router firmware components responsible for accurate usage calculation and limit enforcement, improving system reliability and operational billing accuracy by 22%.
Education
B.Tech in Computer Science and Engineering
National Institute of Technology, Agartala, 2014 — 2018
Contact
Email dhrtklpsh82@gmail.com
Frequently asked questions
Who are you and what do you do?
I am Dhriti Kalpa Saha, a senior full-stack and AI engineer based in Agartala, India. I have more than ten years of experience building AI-enabled enterprise software and high-performance SaaS products, mostly across healthcare, fintech, education and defense. I currently work as a Senior Staff Software Engineer at Appato.
What are you working on right now?
I am a Senior Staff Software Engineer at Appato. My main project is Takshala, an AI-first analytics platform that gives schools natural-language access to their institutional data. I also lead work on financial analytics platforms for hedge fund and prop-trading clients, and on an orchestration protocol for autonomous AI agents.
How many years of experience do you have?
More than ten years. I started in 2015 building web platforms, and have worked across healthcare, fintech, education and defense since.
What is your technical stack or strongest technology?
TypeScript, React and Next.js on the front end; Node.js, Golang, Python and Rust on the back end, plus Solidity for on-chain work. For data I use PostgreSQL, MongoDB, Redis and Neo4j, and for infrastructure Docker, Kubernetes, AWS and GCP. On the AI side the focus is LLM applications, RAG systems, agentic workflows and knowledge graphs.
Do you have AI or machine learning experience?
Yes — it is the core of my recent work. That covers LLM applications, RAG systems, agentic workflows, knowledge graphs, prompt engineering, context orchestration, MLOps and computer vision. I built an air-gapped knowledge-graph RAG platform at DRDO-DMDE that scored 0.86 on the RAGAS scale, and an orchestration protocol using ERC-8004 Trustless Agents.
What is your experience with RAG?
I built an air-gapped, knowledge-graph-based RAG platform at DRDO-DMDE. It covers the whole pipeline — document ingestion, OCR, image parsing, chunking, embedding generation and hybrid retrieval — and runs with no network access at all. It reduced report generation time by 50% and scored 0.86 on the RAGAS scale. This site is a small RAG system too: hybrid retrieval over my own notes.
What industries have you worked in?
Healthcare, fintech, education and defense. In healthcare I have worked on a cancer management and symptom tracking platform, a connected hair-scanning device, and a CMS used across hospitals and clinics in Southeast Asia. In fintech, financial analytics and portfolio dashboards for hedge fund and prop-trading desks.
What is Takshala?
Takshala is an AI-first analytics platform for schools. It gives staff natural-language access to institutional data and generates insight across academic, operational and administrative workflows. I architected it and built the core AI analytics engine; it cut manual reporting and analysis effort by 60%.
Do you have front-end experience?
Yes, and it is where I started. I was a senior frontend engineer at Klinify for four years, leading the frontend team on a healthcare CMS deployed across thousands of hospitals and clinics in Southeast Asia. I modernised a legacy codebase for a 30% gain in development velocity, and optimised core web vitals and observability for a 35% reduction in software-related downtime. React, Next.js and TypeScript are still what I reach for first.
Do you have blockchain or web3 experience?
Yes. At Appato I built an orchestration protocol using ERC-8004 Trustless Agents, with on-chain Identity (ERC-721), Reputation and TEE Validation registries, supporting over 10,000 autonomous agent interactions. Solidity is part of my stack.
Have you led teams?
Yes. I currently lead multidisciplinary engineering teams across AI, fintech, healthcare and enterprise platforms at Appato, with a 95% on-time delivery rate. Earlier, at Klinify, I led the frontend team on a healthcare CMS deployed across thousands of hospitals and clinics.
What is your biggest measurable impact?
A few I would point to: 0.86 on the RAGAS scale for the DRDO-DMDE retrieval platform, a 60% reduction in manual reporting effort with Takshala, over $50M in daily assets supported by the trading platforms I led, and more than 10,000 users on the cancer management platform.
What is your educational background?
I have a B.Tech in Computer Science and Engineering from the National Institute of Technology, Agartala, completed between 2014 and 2018.
Where are you based?
I am based in Agartala, India.
How can I get in touch with you?
Email is the most reliable: dhrtklpsh82@gmail.com. You can also reach me on +91 97749 64682.
Are you open to new roles or freelance work?
I have not put my availability on this site. Email me at dhrtklpsh82@gmail.com with what you have in mind and I will give you a straight answer.
What are your salary expectations?
I have not put numbers on this site. Send the role details to dhrtklpsh82@gmail.com and I will give you a straight answer.
Are you married, how old are you, or what is your family like?
I keep personal details off this site. Happy to talk about anything work-related — projects, skills, or how to get in touch.
