Amrut Jadhav

Technical Lead

• Built a black-box offer underwriting system from scratch on closed infra, extending lending from a limited user subset to PhonePe's entire user base — cut offer generation per user from ~10 days to milliseconds and made lending BCP-ready.
• Embedded compliance and security controls into the black-box offer pipeline, enabling scale without compromising regulatory or security posture.
• Led service-level migration to decouple core offer generation from experimentation, unlocking independent product experiments and cutting go-live TAT for new experiments.
• Built an agentic debugging flow for offer generation, reducing issue triage from ~2 hours to ~40 seconds.
• Building RAG-based agents for on-call alerts to cut false positives, speed up root-causing, and guide new engineers to the right starting point.
• Led adoption of LLM-assisted development across services — structured context for LLMs and MCP integration to reduce token usage.
• Drove cross-functional alignment with Product, Risk, Compliance, and lending partners.
• Java, Dropwizard, Kafka, RabbitMQ, MariaDB, Aerospike, Vector DBs

Technical Lead

• Led DevOps initiatives ensuring system stability, scaling infrastructure, and utilizing K8s for auto-scaling and cost reduction. Led overall system architecture design of the platform.
• Migrated the whole infrastructure to K8s for auto-scaling and cost reduction.
• Established on-call, deployment, and incident-management processes, raising the team to industry-standard engineering practices.
• Achieved a 95% reduction in API response time through multi-threaded architecture and optimized database connectivity. Decreased API issue debugging time by 98% with centralized logging storage and integrated log tracing.
• Elevated system availability from 97.95% to 99.99886%.
• Introduced proactive monitoring and reduced issue detection time by 96.67%.
• Spearheaded security efforts, ensuring regulatory compliance are fulfilled across infrastructure and applications. Implemented a standardized framework for service stabilization. Aligned the whole team to utilize the same for development. Orchestrated end-to-end deployment automation via CI/CD pipelines.
• Nurtured a culture of innovation, continuous learning , and ownership within the team.
• Django, Flask, FastAPI, Dropwizard, React.js, Python, Java, JS, Postgres, MariaDB, Celery with RaabbitMQ, Next.js, AWS, Terraform

Software Engineer

• Responsible for managing core payment microservices which manages millions of transactions per day.
• Responsible for changing service architecture to reduce response time and reduce payment flow complexities.
• Feature architect for 2 major modes of payments. These modes contribute to 90% of total payments happening through the app.
• Responsible for improving system performance and TPS of API's.
• Introduced GitHub action workflows to simplify service deployment procedure and feature testing.
• Contributed to building a RAAS platform for multiple payment modes.
• Responsible for performing system recovery in case of an incident.
• Spring Boot, Java, Kotlin, AWS Aurora DB, TiDB, Redis, AWS, Distributed systems, System Design

Software Engineer

• Building micro-services for an existing monolithic application using a custom framework built on top of spring boot. This includes adding support for the migration of existing services without affecting performance.
• Building event-driven architecture using Kafka as a message broker. This helped to reduce the service delay.
• Creating background jobs using spring batch framework to perform the heavy tasks of migration.
• Modularized the platform and heavily used packages to extract the business logic and core functionality.
• Handling the deployments of docker images of services through API catalog built over the Kong API gateway.
• Migrated an existing Symfony application to Laravel framework. Reduced codebase complexity by reducing layers.
• Adapted eloquent as an ORM in existing application alongside Laravel.
• Java, Spring Boot, Spring Batch, PHP, Node.js, MySQL, PostgreSQL, Kafka, Redis, Docker

Full Stack Engineer

• Developer for Advertise.do, a platform for automation of advertisements across multiple publishers including Adword and Facebook in a team of 5.
• Built multiple ad services on top of Facebook's Graph API and Google's Adwords API.
• Built A/B testing framework for analyzing ad/campaign performance across publishers.
• Built a statistical reporting framework for the platform to provide in-depth insights into campaign performance.
• Maintained advertise reporting data consistency throughout the database to have consistent behaviors across the system.
• Built custom charts from nvd3.js for campaign performance data representation.
• Ruby on Rails, JavaScript, BackboneJs, MarionetteJs, MongoDB, Sidekiq, Redis, Nvd3.js

Consultant

• Lead the migration of existing monolithic property management applications to AWS.
• Reduced feature deployment time from ~5 hours to 5 mins.
• Reduced avg. response time of high traffic pages by almost 90%.
• Re-organized system architecture to sustain high traffic load.
• Laravel, PHP, Bash, JS, MySQL, Docker, AWS

Engineering Intern

• Created a working prototype model for optimized Funnel Analysis for any given event stream.
• Used Neo4j graph database to build relation between user and activities.
• Ruby on Rails, Neo4j

Pune Institute of Computer Technology

Residential junior college, Ahmednagar

HackHenry Tools

• Built a privacy-first suite of 100+ browser-based developer tools to keep PhonePe engineering compliant, letting teams format and decode payloads, tokens and logs without exposing sensitive data to third-party sites that store input.
• Fully client-side processing; inputs never leave the device, no sign-up, preferences stored locally.
• Node.js, Express.js

Doest

• Doest, your intelligent WhatsApp companion, utilizes GPT models to promptly address user inquiries.
• The platform has garnered over 7000 users within 2 months, with continuous growth.
• Nodejs, Postgres, Whats App API

VisualInfo

• Built image attention and attraction analysis tool that check what user will percieve in the first few seconds after seeing image.
• Platform provides 4 reports such as Attention Heatmap, Area of interest, Opacity, Aesthetics.
• Ruby on Rails, Ruby, JavaScript, Backbone.js, Marionette.js, MongoDB, AWS

CardChecker

• Built a platform to simplify the development workflow of companies. The platform continuously listens for updates from the company's Trello board. On every update, it runs a set of predefined rules against a card. If a card doesn't fulfill the standard, it notifies the user on a slack channel.
• It also makes the on-boarding process of newcomers, easy.
• Several companies adopted this system and simplified their process.
• Node.js, Express.js, MongoDB

HivesEstate - End to End Real Estate Chatbot Solution

• Built a whole end-to-end chatbot builder platform that efficiently uses the chatbot as a medium to drive lead generation for real estate clients.
• The platform provides a seamless flow for building a chatbot which only requires 3 steps to deploy the chatbot from scratch.
• Ruby on Rails, Rack, MongoDB, Redis, BackboneJs + MarionetteJs, AWS

The Atom

• Built a personal assistant to improve user productivity by suggesting/initiating actions according to user behavior.
• It looks over user activity when he/she interacts with the device and based upon these activities, it builds personalized usage profile of the user. This profile helps to automate the repetitive tasks of users.
• Ruby, ZeitGeist Ubuntu daemon, Neo4j

DeepLight

• Built a file search engine inspired by the spotlight of macOS for the Linux based systems. It allows a user to search through file systems in natural language. A user can also search on the basis of the content of files, type of files.
• Python, MongoDB, Pynotify

Newsify

• Newsify is a smart news aggregator Mobile/Web service providing users news in less than 60 words.
• Created a back end platform to handle the aw less communication between admin and news writers.
• PHP, Laravel, MySQL, Android

Eshop Markets

• Built online directory of around 1000+ local shops and stores around cities.
• Platform supports user role based design paradigm.
• PHP, MySQL