Rajarshi Chattopadhyay

Software Architect | AI Systems & Business Applications

Education

M.S., Computer Science (Intelligent Systems) — The University of Texas at Dallas

2018 – 2020

Coursework: Data Structures, Big Data, Machine Learning.

B.Tech., Electronics and Communication Engineering — West Bengal University of Technology, India

2014

Thesis: Design and Development of a RISC Processor based on Harvard Architecture.

Selected Positions

Tesla · Dallas, TX

Staff Software Engineer

Feb 2026 – Present
  • Architect AI agents that generate, run, and explain analytical SQL/DAX queries from natural-language questions over enterprise financial data, replacing manual chart and report building.
  • Design secure data access for external applications and agentic workflows: SQL-injection prevention, row-level security, and token-based access.
  • Built hybrid semantic search (vector search plus BM25) to improve search relevance across an internal analytics platform.

Senior Software Development Engineer

Mar 2023 – Feb 2026
  • Sole owner of a visual, self-service data-engineering platform built on Apache Airflow; led it from requirements and architecture through development to MVP (Sep 2024) and launch (Jan 2025). Features include a visual DAG and task builder, test runs, Git integration with conflict resolution, team workspaces, and staging-to-production promotion. Became the standard pipeline-authoring front end for the data-engineering backend.
  • Sole owner of a financial planning and analytics application giving cost-center leaders operating-expense (OPEX) visibility, with loan, lease, and delivery simulators; optimized query performance and resource usage.
  • Built version 1.0 of an enterprise semantic layer and data modeler (SQL query-builder parser), the foundation for a single enterprise data model and business KPI layer.
  • Developed the AI agent for charts and dashboards, workbook life cycle, and datasource editor of an internal analytics platform.
  • Designed, developed, and delivered a regulatory credit trading application that moved a revenue-critical department from spreadsheets to software, with contract and invoice generation, contract-lifecycle integration, and API integrations with other internal applications.
  • Designed, developed, and delivered an application to productionize machine-learning model discovery and documentation, part of the infrastructure that democratizes data-science work across organizations.

Software Engineer

Mar 2021 – Feb 2023
  • Designed, developed, and deployed a company-wide metadata explorer, improving data discovery and lineage tracking from business-process assets to physical database entities.
  • Set up the security architecture and deployment pipelines for cross-domain applications.

Data Engineer Intern — Fremont, CA

Jun 2020 – Nov 2020
  • Developed and automated ETLs for Enterprise Data Warehouse solutions for Sales, Delivery, and Logistics teams in a time-constrained environment using open-source technologies.

Software Engineer Intern, IT — Tesla · Fremont, CA

Sep 2019 – Dec 2019
  • Automated the handling and monitoring of secured infrastructure-access requests, cutting turnaround time by 90%, in support of the company’s digital transformation.

Software Development Engineer Intern — Copart · Dallas, TX

May 2019 – Aug 2019
  • Designed and built a scalable RESTful microservices system for cross-verification and reconciliation of vendor check payments with the bank, preventing fraud worth thousands of dollars each month.
  • Developed and maintained web application modules and REST microservices for billing and payments operations; explored architecture feasibility and trade-offs and prepared design documents.

Software Systems Engineer — IBM · Kolkata, India

Dec 2014 – Dec 2018
  • Developed and supported enterprise systems (client application back end) for a global shipping operation (A.P. Moller-Maersk).
  • Built MQ reporting tools: a Java UI for one-click daily queue-status reports and Bash automation for daily queue monitoring, eliminating manual effort.
  • Provided on-the-job training on application system architecture and databases; received the IBM Best Emerging Techie Award (A.P. Moller-Maersk Diamond Awards, 2016).

Earlier: Administrative Assistant (student worker), Office of Graduate Education, UT Dallas (2018–2019); Embedded Systems Engineer Apprentice, Hi-Q Solutions (2012–2013; FSMs, assembly, VHDL); engineering apprenticeships at Bharat Sanchar Nigam Ltd. and Siemens VAI Metal Technologies (2012–2013).

Selected Projects

AI Agent Stack (2026)

Modular open-source reference for a seven-layer AI agent (APIs, tools, RAG, MCP, plugins, skills, integration); each module is a runnable demo with tests, wired into one traced pipeline. Python. github.com/likarajo/ai-agent-stack

Agent Debugger (2026)

Local Docker debugger that pauses risky agent tool calls (for example email.send) for human Approve or Deny, with a timeline and audit log; pairs with AI Agent Stack. github.com/likarajo/agent-debugger

AI agent for energy data (2026)

Safe natural-language-to-SQL agent over public U.S. EIA electricity-generation data: questions become parameterized single SELECT statements (user text never copied into SQL), a checker blocks writes, comments, extra statements, and out-of-schema columns, and unanswerable questions are refused. Python, SQLite. likarajo.github.io/work/ai-agent-chatbot

Neural Machine Translation

LSTM seq2seq encoder-decoder translation model in Keras. github.com/likarajo/language_translation

COVID-19 Sentiment Analysis (2020)

Sentiment analysis of COVID-19 tweets in R. github.com/likarajo/covid19Sentiments

Customer Targeting

Binary classification model trained with multiple classifiers to predict a customer’s product interest from ad-click and search data. github.com/likarajo/customer_targeting

Sentiment Analysis and Visualization

Spark Streaming application for live tweets via the Twitter API, sentiment analysis with Stanford CoreNLP, and real-time visualization with ZooKeeper, Kafka, Elasticsearch, Logstash, Kibana. github.com/likarajo/twitter_sentiments

More: house price prediction (Spark on AWS EMR), movie-review sentiment (CNN/RNN with GloVe), temperature-anomaly mapping (NASA data), Elasticsearch search app, chatbot. Full list: likarajo.github.io/Projects

Awards

  • IBM Best Emerging Techie Award, A.P. Moller-Maersk Diamond Awards, 2016.
  • First Runner-Up, IETE Technical Paper Contest, 2014: 16-bit RISC processor on an FPGA; also presented at a national seminar on Advanced Microprocessor and VLSI Devices.

Certifications & Badges

Digital badges: credly.com/users/likarajo

  • Oracle Certified Associate, Java SE 8 Programmer (2017)
  • Chegg Skills AI Prompt Engineering Certificate (2024)
  • IBM: Big Data Foundations L1 (2017), L2 (2019); Hadoop Foundations L1 (2017), L2 (2019); Hadoop Programming L1 (2017); Hadoop Data Access L1 (2019); Hadoop Administration L1 (2019); Spark L1 (2019); Docker Essentials (2019); IBM Cloud Kubernetes Service (2019); Blockchain Essentials (2017); Enterprise Design Thinking Practitioner (2016)

Activities & Writing

  • Mentor to interns and new engineers; 2013 talks on computer architecture (IETE seminar) and wireless telecom.
  • HackForHumanity 2018 hackathon; member and volunteer, Graduate Student Assembly and ACM, UT Dallas.
  • AI Agent Stack series (seven modules: APIs, tools, RAG, MCP, plugins, skills, integration) on likarajo.github.io and LinkedIn; technical articles on Text-to-SQL unanswerable-question refusal and data-agent safety; Likarajo Tech YouTube channel (AI Agent Stack end-to-end with Agent Debugger). Earlier writing on LinkedIn and likarajoblogs.wordpress.com (Docker, CI/CD, RISC design, COVID-19 sentiment).