$ whoami

> Aravind Sridhar_

Lead Software Engineer · AI-Forward Full-Stack

AI-forward lead engineer with 9 years in software, the last 5 in fintech. I lead projects end to end across LLM tooling, machine learning, full-stack product, DevOps, and networking, working with product, infrastructure, and leadership to ship them.

↓Download Resume

 \(˘▿˘)/
   | |  
  /| |\  
 (_) (_)

  (\ /)
  (•ω•) ⌚
  o( )o
   (|)

$ follow_the_white_rabbit

> Experience

OneMain Financial

Lead Software EngineerJul 2023 – Present

–Leading an in-app LLM insights platform: a prompt harness that lets product test prompts against multiple data sources (live in dev), plus shadow-run infrastructure that ranks prompt results to decide which AI insights reach customers.
–Improving merchant matching, the engine that links recurring charges to our merchant database and powers user-facing "cancel this" and "negotiate this bill" actions. Better matching has grown the subscription-cancellation and bill-negotiation queue ~1.5x, lifting user engagement and driving the business.
–Led Plaid Asset Report token sharing to show cashflow, bills, and subscriptions before a user links a bank, plus one-click bank connection to cut onboarding friction. User activation went from ~4% to ~6% with Asset Report and to ~12% with one-click, a 3x lift. Owned architecture, frontend and backend code, QA requirements, Datadog performance tracking, and Periscope funnel metrics.
–Led Due Soon, an ML bill-date prediction engine, from whiteboard to production for ~100K bank accounts: forecast accuracy up from ~70% to ~80% and app engagement up ~6%.
–Led the Convox-to-Kubernetes infrastructure migration with zero downtime; cut Datadog spend 30–40% (~$80K/yr), passed a clean SOC 2 audit, and stood up team on-call.
–Integrated OneMain's enterprise domain gateway across software, DevOps, and networking teams, building the Plaid token-sharing GraphQL subgraph (~60K requests/day).
–Led the engineering effort to rebrand Trim to MyMoney (~1M users), authoring the runbook for synchronized execution across application code and DevOps infrastructure.

Senior Software EngineerOct 2021 – Jul 2023

–Directed the technical design and launch of the Instant Advance product line, managing the end-to-end lifecycle from proof of concept to production, including eligibility engine architecture and TabaPay payment integrations.
–Optimized core MyMoney algorithms through regex pattern matching to significantly increase subscription recognition accuracy and drive growth in user cancellation requests.
–Re-engineered negotiation backlog refinement systems to support expansion of medical bill negotiation services, ensuring scalable data processing for new product verticals.
–Developed a conversational SMS tracking system that groups asynchronous text messages into coherent conversations to improve user experience and internal support visibility.

–Build proof-of-concepts to validate technical direction and de-risk architectural decisions
–Tech lead of a 6-engineer team and SME on 4 of its projects; mentored two engineers to promotion
–Drive cross-functional collaboration across product, DevOps, networking, and leadership, and brief the servicing lane's three engineering teams twice a month

Q-Centrix

Software EngineerAug 2019 – Sep 2021

–Built full-stack features on a Ruby on Rails and React clinical data platform
–Automated complex clinical workflows by aggregating data from disparate sources including healthcare integrations (Mirth)
–Improved end-user experience through data-driven UI improvements and performance optimizations

Raytheon

Software Engineer IIJul 2018 – Jul 2019

Software Engineer IJul 2017 – Jul 2018

–Developed real-time embedded software in C on VxWorks for large-scale tracking systems
–Implemented performance-critical algorithms including Kalman filters and track loops
–Built GPS-related systems and native GUI applications in C++ and C# (WPF)

UC San Diego Health

RF Engineering InternJun 2016 – Jun 2017

–Designed, tuned, and tested RF coils for MRI systems in the UCSD MSK research lab
–Collaborated with faculty, senior engineers, and radiologists to evaluate new coil designs
–Gained production experience at GE Healthcare (Aurora, OH) learning RF coil design and manufacturing

  /\_/\  
 ( ≧◡≦) 
  \   /  
  ‿‿‿‿

 ≋≋≋≋≋≋ 
(●)(●)(●)
 ≋≋≋≋≋≋ 
   🐰

$ cd ../??/!!/%%/$$$$/weee/../../projects

> Projects

Insider Stock Watch↗Live site · Private Repo ▶

–Full-stack investing platform that turns every morning's SEC insider filings into a ranked, explained watchlist.
- ↳Daily-refreshed insider buying rankings with per-stock news, interactive charts, and company context from SEC financials and Wikipedia.
- ↳Practice trading: members get $10,000 in virtual money to build a portfolio and learn without risking real cash.
- ↳A stock-basics course that teaches newcomers how to read charts and filings, plus a saved-stock watchlist.
–Live ML trading engine built to beat the S&P 500 (SPY), gold (GLD), and US bonds (AGG).
- ↳12-model LightGBM and random-forest ensemble on 48 point-in-time features, trained on 23.5K insider events since 2008.
- ↳Fully automated daily pipeline scores each night's filings before the market opens and marks long/short paper portfolios to market against all three benchmarks.
- ↳Every strategy has to clear walk-forward validation, an untouched holdout, survivorship and slippage stress tests, and a pre-registered live-trial bar.
–Stack: Python, Node.js, Ruby on Rails, PostgreSQL (Neon), GitHub Actions, Vercel.

AI Learnings↗Inspired by Maddy Zhang ▶

–MCP Exchange Rates: A Model Context Protocol server using exchangerate.host as a real-time data source.
- ↳Exposes a convert_currency MCP tool for real-time currency conversion between any two pairs.
- ↳Includes a standalone web UI for manual testing and validation of tool outputs.

NeetCode Submissions↗

–Python solutions to NeetCode.io coding interview problems, auto-synced from the platform and organized by topic.

$ eat_me --grow

> Grad School

Bayesian Statistics — ISYE 6420↗OMSCS · Private Repo ▶

–Final project: Bayesian gender-based analysis of diabetes risk factors using CDC survey data.
- ↳Investigated how risk factors differentially affect diabetes likelihood in men vs. women, given that 11.6% of Americans (34.6M) were diagnosed as of 2021.
- ↳Modeled posterior distributions over risk factor coefficients to quantify uncertainty in gender-stratified predictions.

Machine Learning for Trading — CS 7646↗OMSCS · Private Repo ▶

–Final project: Built a Random Forest Bag Learner trading strategy and benchmarked it against a rule-based manual strategy.
- ↳Implemented a market simulator supporting BUY (+1000 shares), SELL (−1000 shares), and HOLD positions.
- ↳Compared learner-driven signals against manually crafted technical indicators across in-sample and out-of-sample periods.

$ ask_the_caterpillar

> Skills

./languages/

RubyTypeScriptJavaScriptCC++PythonSQLC#

./ai & ml/

LLM agentsClaude CodeMCPLightGBMscikit-learnpandas

./frameworks/

Ruby on RailsReactNext.jsNode.jsApollo RouterGraphQL

./infrastructure/

AWSKubernetesDockerTerraformPostgreSQLRedisDatadogCI/CD

./tools/

GitLinuxVxWorksMirthFigma

$ curiouser --and-curiouser

> Education

Georgia Institute of Technology·M.S. Computer Science — Machine Learning (part-time)

Expected 2028

UC San Diego·B.S. Computer Engineering

2013 – 2017

  ∿/\_____/\∿ 
 ≋(  ✦   ✦  )≋
 ≈ ( ∞  ‿ ∞ )≈
  ∿)  ~  ~  (∿
  ≋(_________)≋
  ≈  ∿ ∿  ∿  ≈

  ╔══════════╗  
  ║  10 / 6  ║  
  ╚════╤═════╝  
  (◈   ∧   ◈)  
  (  ∧_u_∧  )  
  /|   ~~~  |\ 
 /  |_______|  \

  ≋≋≋≋≋≋≋≋≋≋≋≋≋≋≋≋  
 (●)(●)(●)(●)(●)(●) 
  ≋≋≋≋≋≋≋≋≋≋≋≋≋≋≋≋  
          🐰

∿≋ we're all mad here ≋∿

$ howdy

> Get in touch

Open to new opportunities, collaborations, or just a good conversation.

✉aravindsridhar66@gmail.com