TRUSHAR TANDEL
Stamford,CT
Summary
Enterprise software and AI engineer spanning .NET, Angular, Python, and Google Cloud. Delivers microservices, RAG evaluation frameworks, Document AI integrations, and multi-agent orchestration for classification, extraction, validation, and reconciliation across complex document processing workflows.
Overview
22
22
years of professional experience
Work History
Full Stack Software Developer with AI/MLOps Engineer
ClearDox (Freepoint Commodities LLC)
Stamford, CT
02.2022 - Current
- Supported full stack development for Cleardox enterprise document processing platform for commodity trading across C#/.NET Core backend, Angular frontend, and Google Cloud Platform services.
- Built backend services and microservices in C# and .NET Core using Clean Architecture and modular service design.
- Created low-latency C#/.NET REST services to ingest data securely from Google APIs and AWS APIs, connecting global workflows with internal data pipelines and enterprise systems.
- Developed high-performance internal dashboards with Angular and Angular Material for data science and trading analytics teams.
- Designed multi-agent enterprise document processing system with classification, extraction, validation, and reconciliation agents, using tool calling, MCP-based integrations, RAG, and autonomous agent workflows in Microsoft Agent Framework (C#) built on Semantic Kernel with ChatCompletionAgents, AgentGroupChat, and custom C# tool plugins.
- Built production RAG pipelines in LangChain (Python) and Semantic Kernel (C#) with Google Cloud SQL (PostgreSQL/pgvector), adding custom chunking, embedding generation, semantic retrieval, and evaluation frameworks for accurate, auditable, domain-grounded outputs for commodity trading documents.
- Used Google Cloud SQL for PostgreSQL (pgvector) for secure, low-latency embedding storage behind RAG retrieval, with BigQuery to ground LLM responses in domain-specific enterprise data.
- Designed multi-shot prompting framework with specialized prompt categories for NER, classification, summarization, comparison, reconciliation, question answering, and workflow orchestration, supporting one platform for varied document processing use cases across complex commodity and financial sheets.
- Integrated Gemini LLM into backend through a prompting framework for structured data extraction from complex commodity and financial sheets.
- Integrated Document AI custom extractor and custom classifier models into document processing workflows to parse, segment, categorize, and normalize multi-page corporate documents.
- Established model evaluation and upgrade framework with golden test datasets and parallel Gemini version comparisons (2.5 Flash vs 3.5 Flash) to measure accuracy, completeness, and format compliance before production rollout.
- Implemented RAG evaluation methods with retrieval relevance scoring, answer faithfulness checks, and golden dataset benchmarking to support auditable, traceable AI outputs that meet enterprise compliance standards.
- Automated CI/CD with GitHub Actions to build, test, and deploy services to GCP environments with unit testing in NUnit.
- Reduced manual document review time from 2–3 hours for analysts to seconds through autonomous processing with higher accuracy.
- Established a model evaluation and upgrade framework with golden test datasets and parallel Gemini version comparisons (2.5 Flash vs 3.5 Flash) that measured accuracy, completeness, and format compliance before production rollout.
- Implemented RAG evaluation strategies including retrieval relevance scoring, answer faithfulness checks, and golden dataset benchmarking to ensure auditable, traceable AI outputs meeting enterprise compliance standards.
- DevOps & Impact:
- Automated CI/CD with GitHub Actions to build, test, and deploy services to GCP environments, with unit testing in NUnit.
- Replaced manual document review that took analysts 2–3 hours with autonomous processing that completes in seconds with higher accuracy.
- Environment: C#, .NET Core, Python, Microservices, Clean Architecture, Angular, Angular Material, GCP, Google Cloud SQL (PostgreSQL/pgvector), BigQuery, Vertex AI (Console, Workbench, Pipelines), Document AI, Google APIs, AWS APIs, Microsoft Agent Framework, Semantic Kernel, LangChain, ChatCompletionAgents, AgentGroupChat, C# Tool Plugins, MCP, Gemini LLM, RAG, Custom Chunking & Embedding, Prompt Engineering, GitHub Actions, NUnit.
Senior .NET Developer
Vistra Energy
01.2019 - 02.2022
- Engineered independent, loosely coupled microservices in ASP.NET Core Web API with C#, hosted on Microsoft Service Fabric, forming the backend for energy customer-facing applications.
- Applied Clean Architecture and the CQRS pattern to separate read and write concerns and keep business logic testable and maintainable across services.
- Developed frontend features with Angular 6, Angular Material, and Bootstrap 4 that consume the Web API layer, backed by SQL Server data stores.
- Built secure authentication, authorization, and data validation middleware, enhancing security for complex energy client accounts.
- Implemented SignalR for real-time communication between backend services and client applications.
- Led enterprise-wide data migration efforts, ensuring accurate data mapping, integrity cleansing, and efficient pipeline planning into production SQL Server instances.
- Wrote unit tests with NUnit and delivered builds through TeamCity and Octopus Deploy CI/CD pipelines, with source control on Bitbucket and Git.
- Provided post-production support, resolving bugs and performance issues to maintain optimal application performance.
- Environment: .NET Core 3.0, C#, Clean Architecture, Microservices, CQRS, SignalR, SQL Server, Angular 6, Web API, Angular Material, Bootstrap 4, AutoMapper, Microsoft Service Fabric, Bitbucket, Git, NUnit, TeamCity, Octopus, Confluence.
Senior .NET Developer
North American Power
10.2016 - 12.2018
- Led the modernization of the legacy front end from AngularJS 1.6 to Angular 6, restructuring the application around components, reusable pipes and services, and reactive forms.
- Built UI with Angular Material and Bootstrap 4, delivering a full stack solution from Web API and database layer through to the Angular client.
- Developed ASP.NET Core Web APIs in C# for seamless integration across web, mobile, and tablet platforms.
- Used Entity Framework as the repository layer over SQL Server, implementing LINQ queries for CRUD operations.
- Oversaw continuous integration processes using TeamCity and Octopus Deploy; created NUnit test projects and performed code analysis with ReSharper.
- Environment: .NET Core 2.0, C#, SQL Server, Entity Framework, AngularJS, Angular 6, Web API, Angular Material, Bootstrap 4, Bitbucket, Git, NUnit, TeamCity, Octopus, ReSharper, Rally
Lead Developer
World Bank
03.2016 - 09.2016
- Developed Web API services with Entity Framework and Microsoft SQL Server, integrating the frontend using jQuery and HTML5.
- Managed client requirements and technical delivery for an ASP.NET MVC 5 portfolio tracking application built with C# and Telerik Kendo UI.
- Owned code deployment and production support during integration; guided the team and coordinated daily with offshore resources.
- Environment: Visual Studio 2013, C#.NET, ASP.NET MVC 5, Web API, Telerik Kendo, jQuery, Entity Framework, HTML5, TFS, MS SQL Server.
Lead Developer / Onsite Coordinator
Citi Bank
07.2012 - 03.2016
- Architected a three-tier enterprise banking system using C#/ .NET with ADO.NET data access.
- Designed analytical stored procedures and views in SQL Server and authored dynamic T-SQL queries for complex data operations.
- Served as Team Lead and Onsite Coordinator for requirement analysis, design, development, testing, documentation, and production deployment.
Lead Developer
Lloyds Bank
10.2009 - 10.2012
- Designed Advanced-IRB LGD and EAD modeling platforms under Basel II using optimized C# Web Services, and developed the Business Layer of the CARE (Corporate Markets Advanced Ratings Engine) application.
- Wrote unit tests with NUnit and improved application throughput through stored procedure optimization.
Consultant
ICICI Bank
01.2008 - 01.2009
- Additional Experience: Consultant at ICICI Bank (2008–2009) | Developer at Kaushalam Technologies (2005–2007)
Developer
Kaushalam Technologies
01.2005 - 01.2007
- Additional Experience: Consultant at ICICI Bank (2008–2009) | Developer at Kaushalam Technologies (2005–2007)
Education
Master of Computer Applications (MCA) -
B.Sc. - Computer Science
Skills
- C#
- NET Core
- ASPNET Core Web API
- Backend Development: NET Core, ASPNET Core Web API, ASPNET MVC, RESTful APIs, Microservices, Microsoft Service Fabric, SignalR, Automapper, Clean Architecture, CQRS
- Frontend Development: Angular 6, AngularJS, Angular Material, Reactive Forms, Components, Pipes, Services, Bootstrap 4, Kendo UI, jQuery
- Databases & Data Access: Microsoft SQL Server, PostgreSQL (pgvector), Google Cloud SQL, BigQuery, Oracle, Entity Framework, LINQ, ADONET, T-SQL, Stored Procedures
- Cloud Platforms: Google Cloud Platform (GCP), Vertex AI (Console, Workbench, Pipelines), Document AI, Google Cloud SQL, BigQuery Azure (limited experience): Microsoft Certified: Azure Fundamentals (AZ-900)
- AI & ML Integration: Microsoft Agent Framework, Semantic Kernel, ChatCompletionAgents, AgentGroupChat, C# Tool Plugins, Model Context Protocol (MCP), Gemini LLM, Retrieval-Augmented Generation (RAG), LangChain, Custom Chunking & Embedding Strategies, RAG Evaluation Frameworks, Multi-Shot Prompt Engineering, Python AI/ML Integration
- Architecture & Engineering: Clean Architecture, CQRS, Three-Tier Architecture, REST API Design, Enterprise Application Integration, Automated Testing, Performance Optimization, Application Modernization
- DevOps & CI/CD: GitHub Actions, TeamCity, Octopus Deploy, Bitbucket, Git, TFS, NUnit, ReSharper, Rally, Confluence
- Languages: C#, Python, T-SQL, JavaScript, TypeScript, HTML5, CSS
- Google Cloud Platform
- Microsoft Agent Framework
- Semantic Kernel
- AI & ML Integration: Microsoft Agent Framework, Semantic Kernel, ChatCompletionAgents, AgentGroupChat, C# Tool Plugins, Model Context Protocol (MCP), Gemini LLM, Retrieval-Augmented Generation (RAG), LangChain, Custom Chunking & Embedding Strategies, RAG Evaluation Frameworks, Multi-Shot Prompt Engineering, Python AI/ML Integration
Personal Information
Title: Senior .NET AI Developer | Technical Lead
Timeline
Full Stack Software Developer with AI/MLOps Engineer
ClearDox (Freepoint Commodities LLC)
02.2022 - Current
Senior .NET Developer
Vistra Energy
01.2019 - 02.2022
Senior .NET Developer
North American Power
10.2016 - 12.2018
Lead Developer
World Bank
03.2016 - 09.2016
Lead Developer / Onsite Coordinator
Citi Bank
07.2012 - 03.2016
Lead Developer
Lloyds Bank
10.2009 - 10.2012
Consultant
ICICI Bank
01.2008 - 01.2009
Developer
Kaushalam Technologies
01.2005 - 01.2007