Hi, my name is
Senior backend engineer with seven years of experience building, scaling, and operating distributed-systems infrastructure at hyperscale. Deep expertise across Python, C++, Rust, and Java, with working fluency in Go and PHP. Recognized for end-to-end ownership of complex subsystems — design, implementation, rollout, on-call, documentation — and for raising the bar through technical writing, recorded tech talks, and onboarding curricula.
Started as an Application Developer at JPMorgan Chase & Co. in Mumbai (2014–2017), interned at Meta in 2018, and joined Meta full-time in 2019. Along the way co-authored published research on neuro-fuzzy / simulated-annealing optimization and stock-market prediction with neural networks, contributed engineering volunteer work to Pratham Education Foundation, and earned an ML certification from Stanford and a Python specialization from the University of Michigan.
Hello! I’m Chirag — a backend engineer at Meta working on Core Systems. I’m a versatile and productive team player with passion and 7+ years of experience in distributed computing, backend technology, and data analytics. I communicate effectively, collaborate with stakeholders while identifying unmet needs that aren’t obvious, and I’m result-driven.
I’ve developed new solutions and enhanced existing processes in technologies including ZooKeeper, Linux, C++, Python, Rust, Java, Hack, Shell scripting, MySQL, HP-Vertica, ETL tooling, and a wide range of web technologies. During my time at JPMorgan Chase I performed production release management and gained working knowledge of the lifecycle of the mortgage-backed securities market.
My interests center on distributed computing, software development, data analytics, and machine learning. Outside of code I’m a tech enthusiast and creative tinkerer — and a strong believer that the best engineering happens in teams. Here are a few technologies I’ve been working with recently:
// recently working with
Master of Science · Computer Science
Jan 2017 — Dec 2018 · Raleigh, NC, USA
// courseworkAdvanced Computer Networks · Advanced Database Management Systems · Analysis of Algorithm and Design · Data Driven Decision Making · Data Warehouse and Mining · Database Management Systems · Machine Learning · Spatial and Temporal Data Mining · Web Engineering · Robotics & Artificial Intelligence · Operating Systems
Bachelor of Engineering (B.E.) · Computer Engineering · University of Mumbai
Jan 2010 — Dec 2014 · Mumbai, India
5th rank overall amongst all students from various specializations in the college. Authored three IJCSE research publications during this period.
// courseworkObject-Oriented Design and Development · Object-Oriented Software Engineering · Data Structures and Files · Computer Programming · Computer Graphics · Mathematics · Communication Skills
Featured project
Microdata API for the Mozilla web-rendering engine, October 2017.
Contributed to the Servo open-source web-rendering project (Mozilla) by designing an algorithm to extract machine-readable metadata from web content. The metadata is used to generate vCard / vEvent files via servoshell embedding.
The contribution adds a Microdata API surface to Servo’s DOM, enabling embedded consumers to programmatically extract structured data from rendered HTML pages — a small but pivotal piece of plumbing for the web-platform interop story.
── tech ─────
Featured project
Spatial-temporal car-pooling marketplace deployed on AWS Elastic Beanstalk.
Managed and designed an Amazon Elastic Beanstalk PaaS deployment for a responsive web app built with Node.js and MongoDB. Wolfpool facilitates planning and sharing of car rentals, Uber, and Lyft rides via a spatial-temporal rule-based system — useful where UberPool / Lyft Line are unavailable.
The system matches users by route similarity within configurable time-windows, then offers a shared booking pathway. Elastic Beanstalk provided the rolling-deploy story; MongoDB held the user/route store with geo-indexed lookups.
── tech ─────
Featured project
Statistical and sentiment analysis on tweets with custom indices and Tableau visualizations.
Statistical and sentiment analysis on tweets using Python NLTK to formulate custom indices and scores. Investigated whether extracting product/service reviews from Twitter could be automated end-to-end, given the noise and brevity of the platform.
Tableau visualizations exposed the contributing factors — sentiment polarity, hashtag co-occurrence, temporal clustering — and a Cassandra/Spark pipeline ran the ingestion and aggregation at the volume needed to be statistically meaningful.
── tech ─────
Featured project
Cloud-based HR & expense system for an NGO. Projected savings ~$150,000/year.
Used Joget Workflow Tool, Java, and Microsoft Azure to build a responsive web/mobile platform for Pratham Education Foundation. Automated manual tasks, scaled operations — projected savings of ~$150,000/year. Co-implemented a Selenium test-suite for sustainable downstream development.
The platform replaced spreadsheet-driven HR + expense workflows for a multi-state NGO. Joget’s low-code workflow primitives kept the maintenance footprint small enough for a volunteer-led engineering rotation to operate going forward.
── tech ─────
// Graduate (NCSU 2017–2018)
Recommender system for automobiles based on indirect user profile, car specifications, and semantic analysis of reviews and news available for each car, using NLP.
Analyzed factors that influence house prices using regression and other models; established correlations between spatial-temporal influence of crime rates and pricing.
Contributed to the Servo open-source web-rendering project (Mozilla) by designing an algorithm to extract machine-readable metadata from web content for vCard / vEvent generation.
// Undergraduate & Early (2012–2017)
Hybrid approach combining Adaptive Neuro-Fuzzy Inference System (ANFIS) with Simulated Annealing to determine paths based on distance, profit, and probability of deal. Foundation for two IJCSE publications.
Mobile-responsive web platform for a logistics marketplace covering all transportation needs. Used Google Maps API for distance computation; auto-generated email + SMS reminders for transactions.
Real-time, paperless attendance tracker for Teach For India, built in under 24 hours at JPMorgan Code For Good Challenge Mumbai 2013. IVRS, SMS, lightweight mobile app. Winning team.
Desktop application that simulates a mobile-user position in a GSM environment. Visualized frequency handover, call drop, roaming. Graph-coloring algorithm for base-station assignment; dynamic frequency allocation/reuse. HLR/VLR maintained.
Desktop application for timed auto-shutdown of computers — useful for large lab setups where dozens of machines need to come down on a schedule.
Data-analysis tool extracting from .sql/.accdb/.txt/.xlsx, cleaning data, loading into MS Access. SQL search query engine; implementations of Naïve Bayes, K-Nearest Neighbours, K-Means and Hierarchical Clustering.
Designed/developed company website with offline-readable content and a dynamic admin panel. Optimized loading using a hidden inner-HTML division technique.
Online platform for engineering students/faculty to publish research. Admin panel for paper uploads and metadata management.
Platform addressing the programming needs of students. Designed layouts, database, navigation. Full-text search on metadata for fast lookup.
2D Java-applet game where the goal is to destroy the opponent’s castle using cannons. Projectile motion impacted by wind direction/speed; impact accuracy computes damage.
International Journal of Computer Science and Engineering (IJCSE)
ISSN (Online): 2278-9979 · ISSN (Print): 2278-9960
Jul 30, 2014
Implemented a hybrid approach comprising an Adaptive Neuro-Fuzzy Inference System (ANFIS) and Simulated Annealing algorithm to determine paths based on distance and other factors like profit and probability of deal.
@article{jain2014tsp_impl,
title = {Optimization of Traveling Salesman Problem for an Enterprise Resource Planning System | Implementation},
author = {Jain, Chirag and Shah, Richa},
journal = {International Journal of Computer Science and Engineering (IJCSE)},
year = {2014},
month = {jul},
issn = {2278-9979},
url = {http://www.iaset.us/view_archives.php?year=2014_14_2&id=14&jtype=2&page=5}
}
International Journal of Computer Science and Engineering (IJCSE)
ISSN (Online): 2278-9979 · ISSN (Print): 2278-9960
Nov 30, 2013
Proposed the use of neural networks to interpret complex stock-market patterns for traders, enabling quick, calculated, and informed decisions.
@article{jain2013nn_stocks,
title = {Neural Network Applications in Stock Market Predictions},
author = {Jain, Chirag and Jani, Himali and Jain, Shreyansh and Bhowmick, Kiran},
journal = {International Journal of Computer Science and Engineering (IJCSE)},
year = {2013},
month = {nov},
issn = {2278-9979},
url = {http://www.iaset.us/view_archives.php?year=2013&id=14&jtype=2&page=6}
}
International Journal of Computer Science and Engineering (IJCSE)
ISSN (Online): 2278-9979 · ISSN (Print): 2278-9960
Oct 30, 2013
Proposed a practical use case of the classical Traveling Salesman Problem to determine the shortest path, using ANFIS and Simulated Annealing.
@article{jain2013tsp_proposal,
title = {Optimization of Traveling Salesman Problem for an Enterprise Resource Planning System},
author = {Jain, Chirag and Shah, Richa},
journal = {International Journal of Computer Science and Engineering (IJCSE)},
year = {2013},
month = {oct},
issn = {2278-9979},
url = {http://www.iaset.us/view_archives.php?year=2013&id=14&jtype=2&page=5}
}
Python (expert), C++ (expert), Rust (advanced), Java (advanced), Go (working), PHP / Hack (working), C# (working), JavaScript / jQuery (working). Strong typed-IDL design.
Distributed configuration management, distributed computing at hyperscale, RPC frameworks, distributed graph stores, distributed key-value stores, distributed log-store / streaming pipelines (at-least-once semantics), distributed coordination (ZooKeeper / leader election), object storage with atomic uploads and integrity hashing.
Real-time analytics platforms, time-series monitoring, distributed batch pipelines, SQL-on-Hadoop / distributed SQL engines (HP-Vertica for columnar analytics), in-house sampling profilers. ETL with Informatica, BIDS, SQL Server Management Studio. Tableau, NLTK, scikit-learn, Cassandra, Apache Spark.
Container orchestration, service discovery and RPC routing, continuous-deployment pipelines, dynamic feature flags, distributed version control (Sapling / Mercurial / Git fluency). Production release management.
AWS (Elastic Beanstalk), Microsoft Azure, Google Maps API, Node.js, MongoDB, MySQL, MS-SQL. Web stack: HTML, CSS, Bootstrap, jQuery, React, Native Templates, PHP/Hack.
Classification, clustering, neural networks, neuro-fuzzy systems, regression, sentiment analysis, NLP. Hands-on with scikit-learn, NLTK, Tableau, QGIS. Stanford ML certified.
Reliability and on-call, performance optimization, large-scale migrations, cross-language API design and FFI plumbing, production-incident response, static analysis / codemod-driven refactors, build & release infrastructure.
Technical writing, live tech talks, onboarding curriculum design, mentorship, operator-facing CLIs and runbooks, breakglass-procedure authoring. Three published research articles.
JPMorgan Chase & Co. selective early-career rotational program.
JPMorgan Chase, Mumbai 2013. Built a paperless real-time attendance tracker for Teach For India in under 24 hours.
India · undergraduate research presentation.
Dwarkadas J. Sanghvi College of Engineering, across all specializations.
Merit-based undergraduate scholarship.
Qualified · Indian national physics olympiad screening.
Through every layer: design → IDL → service → client → rollout → on-call → docs.
Finished migrations, with the legacy code path removed (never just deprecated).
Observable, reversible, gradient-rolled changes. Mitigate, clean up, then add monitoring.
Documentation is a first-class artifact, not a footnote.
Many small instrumented wins beat one heroic rewrite.
CLIs, runbooks, and breakglass procedures for the engineer who comes next.
Coherent abstractions across language boundaries, with FFI plumbing in between.
When a rollout has to revert, it gets re-attempted, not abandoned.
Joined Meta full-time on Core Systems. Built a multi-axis admission-control service from scratch. Landed in a fleet-wide usage-tracking pipeline. Joined the on-call rotation. First two recorded internal tech talks delivered.
Drove a Python 2 → 3 migration of ~30 modules; owned a release-controller notification stack. Two more recorded tech talks. Began co-authoring onboarding curricula. Cross-pod rotation through a client-API team (Python / C++ / web platform).
Multiple on-call rotations. Contributed to canary/safety subsystems. Led a 5-phase rollout in a major consumer-product surface (with full revert + re-rollout). Drove a series of large codemods. Began long-running graph/expansion-rule cleanup. Wrote a breakglass runbook for an emergency config-update path.
Built and rolled out a new control-plane service from boilerplate to production. Repeated the playbook for a sister service weeks later. Ran a sustained 50+ patch dependency-resolver optimization campaign. Led a snapshot-format migration across all production shards. Authored the canonical IDE setup wiki for the dev fleet. Began driving a weekly design-review series.
Drives design-review series, operational stability, and platform-wide migrations. Mentors new hires through onboarding docs and 1:1 walkthroughs. Active adoption of AI-driven tooling for repetitive maintenance work.
I’m always interested in hearing about distributed-systems problems, infrastructure-platform work, technical mentorship, and ambitious side-projects. My inbox is open — whether you have a question, an opportunity, or just want to say hi.
I’ll do my best to get back to you.
Say hello