Poorvaditya Behre profile photo 1 on Tal

Poorvaditya Behre

SWE @Confido Health • AI x Distributed Systems • Worked in Healthcare, E-commerce

Bengaluru

3+ years of work experience

Confido Health
Nov 202511 mosPresent

United States

CORE PLATFORM • Designed the data & action layer AI voice agents call to run appointment scheduling, prior-auth, payments, refills, patient management & human handoff turning conversation into EHR-backed workflows. • Built a multi-tenant federated platform with per-client environments, enabling forward-deployed engineers to ship bespoke workflows quickly without impacting the shared platform. EHR INTEGRATION • Owned the two-way EHR integration platform for appointments, patients, etc workflows through a unified, capability-driven interface across 30+ EHR systems. • Built a two-tier rate-limit stealing mechanism over a distributed GCRA limiter so live traffic borrows unused background capacity yet never starves under load. APPOINTMENT MANAGER • Architected a high-throughput, DST-safe scheduling engine with multi-tier rules and an atomic Redis/Lua slot-reservation system preventing double-booking under concurrent agent calls. • Designed a cascade search patient-matching algorithm combining deterministic fuzzy and phonetic lookups with LLM-based resolution & configurable per-field confidence thresholds to resolve ambiguous identities across large multi-tenant datasets. • Led the zero-downtime migration of the legacy platform to Go via dual writes, parallel sync, traffic mirroring & per-org cutover with instant rollback. EVENT-DRIVEN • Architected multi-tenant Temporal sync workflows with resumable cursor-based processing and prioritised replay, allowing failed large-range syncs to resume from the last completed batch instead of restarting. • Replaced scheduled polling with a Kafka-based event-driven sync architecture using partitioned topics, DLQs, and self-healing deletion reconciliation across pull and push pipelines. GATEWAY, SECURITY & COMPLIANCE • Built platform security and compliance infrastructure, including custom Kong plugins for routing/authentication and a HIPAA audit service with AES-256-GCM encryption and partition-based PostgreSQL retention.
XCaliber Health
Feb 202511 mosDec 2025

United States

• Developed scalable data product pipelines and orchestration workflows for multi-EHR integrations (athena, elation, ecw, pointclickcare) in a multi-tenant platform. Implemented FHIR interopability based transformation, enrichment and ingestion. • Designed and led migration of existing cloud-native microservices to a hybrid multi-tenant architecture, introducing modular workflows for OnDemand, DataSync, and advanced pre-Lookup functionalities. Reduce redundant EHR API calls by 30%, enhanced system reliability and scalability, and improve query performance by 10%. Tech Stack: Golang, Python, Temporal, Proto, Redis, Postgres, AWS
Semaai
May 202410 mosFeb 2025

Indonesia

Fueled by Challenges, Driven by Growth :) Features rolled out : • Designed and implemented an extension service for the Buy Now Pay Later (BNPL) system, enabling seamless repayment of loans by providing due date extensions. • Integrated Indomaret & Alfamart as new payment providers, expanding offline payment options. • Developed a postpaid payment mechanism via Virtual Account Transfer, enabling upfront payments. • Led the refactoring of the Payment Service, improving scalability and maintainability by adopting Factory and Strategy Patterns, enabling efficient onboarding of new payment providers with minimum setup configurations. • Implemented loan repayment feature for BuyNowPayLater(BNPL) service allowing users to select multiple BNPL loans (fully or partially) for repayment using Open Virtual Account (VA) Bank functionalities, increasing flexibility and ensuring timely payments. Day to Day Responsibilities: • Responsibility doesn’t limit to features only. I solved more than 50+ P0 and P1 bugs with minimum downtime to make the overall system stable and consistent. This helped me to enhanced my debugging skills and problem solving skills in depth. • Actively contributed/discussed to high-level and low-level design decisions, ensuring robust and scalable architecture for business requirements. Tech Stack: Python, Odoo, PostgreSQL, Redis, AWS (Lambda, S3, SQS), Jenkins, Grafana.
Noccarc
May 20231 yr 1 moMay 2024

Pune District, Maharashtra, India

A Journey Well Taken :) Noccarc - a medical device company at the forefront of healthcare innovation. Noccarc was recognised as LinkedIn Top Startups 2024 in the Pune. I was a part of the PEG Software Team, which was responsible for developing solutions and architecting various software systems for v730i (smart ventilator). • Implemented features like product PDF report generation and dealer onboarding workflows in Python, Django, and Django REST for the service app. Initially it was non-production project, I took it as a challenge and made the app production-ready by resolving existing bugs and introducing new features, significantly improving the application. This reduced the onboarding and reporting time significantly. • Designed and developed an over-the-air update system for ventilator v730i using Python, boto3, and ZeroMQ, enabling remote update retrieval and installation. This solution optimized system performance, reducing CPU usage by 5%. • Contributed significantly to the development of a robust device management platform backend, responsible for monitoring, updating, release management of ventilator v730i using Node.js, Redis, and PostgreSQL. • Designed and developed a high-performance data communication system between the UI and embedded backend for the Patient Monitor using C++ 17 and ZeroMQ, capable of handling data at millisecond intervals. Implemented this as a thread-safe independent library, allowing for versatile use in sending and receiving data across various applications. • Developed HL7 interfacing engine module for converting patient vitals observations from ventilator into standard HL7 mapping. • Contributed in the development of Bash Scripts for updating OS features and WiFi, IP logging modules using nmcli, enhancing the connectivity and automation capabilities of the systems.

Poorvaditya’s AI Stack

claude code
claude code
cursor
cursor
hugging face
hugging face
langfuse
langfuse
langgraph
langgraph
livekit
livekit

Education

Bachelor of Engineering - BE, Information Technology

International Institute of Information Technology, Pune

20204 yrs2024

Shri Balaji Public School - Jabalpur, India

20191 yr2020

Join Poorvaditya and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Poorvaditya Behre on Tal

Headline

SWE @Confido Health • AI x Distributed Systems • Worked in Healthcare, E-commerce

About

Hello 👋 I’m Poorvaditya Behre, an experienced software developer who loves solving problems, building, and scaling applications. If you’re a tech geek looking to tackle complex, interesting challenges and are passionate about products, let’s connect! mail : poorvaditya18@gmail.com Technical Skills & Tools: Language : Python, Golang, C++, JS Database : MySQL, PostgreSQL , MongoDB, Redis, Singlestore Message Broker : Kafka, RabbitMQ, ZeroMQ ORM : Seqeulize, Mongoose Api Protocols: REST, gRPC Cloud Services : GCP,AWS Containers : Docker, Kubernetes Frameworks : Gin, FastAPI, Node.js, Nest.js, Django, React, Odoo. Orchestrators: Temporal Library : LangChain, LangGraph, HuggingFace, Pandas , Numpy, Matplotlib. If you can point out which of these are pokemon, let's talk. buymeacoffee.com/poorvaditya

Location

Bengaluru

AI stack

claude code

cursor

hugging face

langfuse

langgraph

livekit

Work experience

Software Development Engineer at Confido Health, Nov 2025 - Present · 11 mos, United States, CORE PLATFORM • Designed the data & action layer AI voice agents call to run appointment scheduling, prior-auth, payments, refills, patient management & human handoff turning conversation into EHR-backed workflows. • Built a multi-tenant federated platform with per-client environments, enabling forward-deployed engineers to ship bespoke workflows quickly without impacting the shared platform. EHR INTEGRATION • Owned the two-way EHR integration platform for appointments, patients, etc workflows through a unified, capability-driven interface across 30+ EHR systems. • Built a two-tier rate-limit stealing mechanism over a distributed GCRA limiter so live traffic borrows unused background capacity yet never starves under load. APPOINTMENT MANAGER • Architected a high-throughput, DST-safe scheduling engine with multi-tier rules and an atomic Redis/Lua slot-reservation system preventing double-booking under concurrent agent calls. • Designed a cascade search patient-matching algorithm combining deterministic fuzzy and phonetic lookups with LLM-based resolution & configurable per-field confidence thresholds to resolve ambiguous identities across large multi-tenant datasets. • Led the zero-downtime migration of the legacy platform to Go via dual writes, parallel sync, traffic mirroring & per-org cutover with instant rollback. EVENT-DRIVEN • Architected multi-tenant Temporal sync workflows with resumable cursor-based processing and prioritised replay, allowing failed large-range syncs to resume from the last completed batch instead of restarting. • Replaced scheduled polling with a Kafka-based event-driven sync architecture using partitioned topics, DLQs, and self-healing deletion reconciliation across pull and push pipelines. GATEWAY, SECURITY & COMPLIANCE • Built platform security and compliance infrastructure, including custom Kong plugins for routing/authentication and a HIPAA audit service with AES-256-GCM encryption and partition-based PostgreSQL retention.

Software Engineer at XCaliber Health, Feb 2025 - Dec 2025 · 11 mos, United States, • Developed scalable data product pipelines and orchestration workflows for multi-EHR integrations (athena, elation, ecw, pointclickcare) in a multi-tenant platform. Implemented FHIR interopability based transformation, enrichment and ingestion. • Designed and led migration of existing cloud-native microservices to a hybrid multi-tenant architecture, introducing modular workflows for OnDemand, DataSync, and advanced pre-Lookup functionalities. Reduce redundant EHR API calls by 30%, enhanced system reliability and scalability, and improve query performance by 10%. Tech Stack: Golang, Python, Temporal, Proto, Redis, Postgres, AWS

SDE1 at Semaai, May 2024 - Feb 2025 · 10 mos, Indonesia, Fueled by Challenges, Driven by Growth :) Features rolled out : • Designed and implemented an extension service for the Buy Now Pay Later (BNPL) system, enabling seamless repayment of loans by providing due date extensions. • Integrated Indomaret & Alfamart as new payment providers, expanding offline payment options. • Developed a postpaid payment mechanism via Virtual Account Transfer, enabling upfront payments. • Led the refactoring of the Payment Service, improving scalability and maintainability by adopting Factory and Strategy Patterns, enabling efficient onboarding of new payment providers with minimum setup configurations. • Implemented loan repayment feature for BuyNowPayLater(BNPL) service allowing users to select multiple BNPL loans (fully or partially) for repayment using Open Virtual Account (VA) Bank functionalities, increasing flexibility and ensuring timely payments. Day to Day Responsibilities: • Responsibility doesn’t limit to features only. I solved more than 50+ P0 and P1 bugs with minimum downtime to make the overall system stable and consistent. This helped me to enhanced my debugging skills and problem solving skills in depth. • Actively contributed/discussed to high-level and low-level design decisions, ensuring robust and scalable architecture for business requirements. Tech Stack: Python, Odoo, PostgreSQL, Redis, AWS (Lambda, S3, SQS), Jenkins, Grafana.

Software Development Engineer at Noccarc, May 2023 - May 2024 · 1 yr 1 mo, Pune District, Maharashtra, India, A Journey Well Taken :) Noccarc - a medical device company at the forefront of healthcare innovation. Noccarc was recognised as LinkedIn Top Startups 2024 in the Pune. I was a part of the PEG Software Team, which was responsible for developing solutions and architecting various software systems for v730i (smart ventilator). • Implemented features like product PDF report generation and dealer onboarding workflows in Python, Django, and Django REST for the service app. Initially it was non-production project, I took it as a challenge and made the app production-ready by resolving existing bugs and introducing new features, significantly improving the application. This reduced the onboarding and reporting time significantly. • Designed and developed an over-the-air update system for ventilator v730i using Python, boto3, and ZeroMQ, enabling remote update retrieval and installation. This solution optimized system performance, reducing CPU usage by 5%. • Contributed significantly to the development of a robust device management platform backend, responsible for monitoring, updating, release management of ventilator v730i using Node.js, Redis, and PostgreSQL. • Designed and developed a high-performance data communication system between the UI and embedded backend for the Patient Monitor using C++ 17 and ZeroMQ, capable of handling data at millisecond intervals. Implemented this as a thread-safe independent library, allowing for versatile use in sending and receiving data across various applications. • Developed HL7 interfacing engine module for converting patient vitals observations from ventilator into standard HL7 mapping. • Contributed in the development of Bash Scripts for updating OS features and WiFi, IP logging modules using nmcli, enhancing the connectivity and automation capabilities of the systems.

Education

Bachelor of Engineering - BE, Information Technology, International Institute of Information Technology, Pune

Shri Balaji Public School - Jabalpur, India

Public links

github: https://github.com/poorvaditya18

leetcode: https://leetcode.com/ironman18