Surya KapilaExperienceProjectsSkillsResume MCPContactPDF

Backend Engineer • Distributed Systems • AI Tooling

Surya Prakash Rao Kapila

I build and own production backend systems that integrate enterprise software at scale, serving 40+ enterprise clients and processing over 100,000 requests a day. My work spans across legacy .NET middleware and modern TypeScript services, evolving business-critical systems without disrupting production while designing reliable integrations between them.

Alongside production engineering, I build AI-powered developer tooling, including RAG systems over production codebases, MCP servers, and autonomous coding agents. I enjoy applying AI where it solves real engineering problems, making complex systems easier to understand, maintain, and evolve.

5+
Years engineering
Across software & systems engineering
40+
Enterprise clients
Serving enterprise platforms in production
100k
Requests / day
Handled with reliability at scale
AI-Native
RAG · agents · pipelines
AI-powered tooling that ships
Building scalable backend platforms. Open to Backend Engineer and SDE II/III opportunities.surya.kapila@gmail.com
01
Experience
May 2025 – Present
Hyderabad, India

Software Engineer

Cadient Talent

  • Primary technical owner of a platform serving 40+ enterprise clients at over 100,000 requests per day, leading production support, client onboarding, and vendor integrations.
  • Develop and maintain a .NET Framework (C#) middleware layer of WCF and ASMX SOAP services and REST endpoints, sitting between the internal applicant tracking system and the external HRIS ecosystem and responsible both for consuming partner integrations and for the public APIs exposed to external consumers.
  • Shipped 6 new integrations and modernised 5 existing ones in 14 months, each run from requirements gathering through interface design, C# implementation and production rollout: ZipRecruiter Jobs, AppCast job board export, UKG Pro new hire export, AccuSource background screening, QJumpers candidate import, and the internal Job, Location and User APIs.
  • Own data integrity at system boundaries, validating inbound and outbound payloads, preventing duplication and loss, and absorbing vendor API changes without client-facing downtime.
  • Resolve customer-reported production issues end to end, tracing failures across the C# layer, SQL Server and third-party APIs, and communicating root cause and remediation directly to enterprise clients and internal stakeholders.
  • Develop and maintain a multi-tenant Node.js and TypeScript service handling Job, Employee and Location synchronisation, with REST APIs on TypeScript, Zod, SQL Server and TypeORM plus retry and reconciliation handling so partial failures do not corrupt client data.
  • Write unit tests for C# services with MSTest, using AI to generate first-pass coverage and reviewing every case before merge.
  • Built internal engineering tooling on Claude Code and a RAG index over the codebase, cutting the time to trace a defect through an unfamiliar integration and shortening onboarding for new joiners.
  • Work through pull requests, peer code review and CI/CD pipelines, holding AI-generated changes to the same review bar as hand-written code.
C# / .NETSQL ServerTypeScript · TypeORM · ZodClaude Code · RAGMSTest
Jul 2023 – Apr 2025
Hyderabad, India

Software Engineer & Consultant

CoSchool

  • Built production REST APIs and backend services in Node.js, NestJS and TypeScript on an Agile team, applying object-oriented design and established service patterns.
  • Designed an event-driven notification platform on Kafka, decoupling distributed services through asynchronous messaging instead of synchronous calls.
  • Cut database load 35% by introducing Redis caching on hot read paths.
  • Improved backend response times roughly 20% through query optimisation and service refactoring.
  • Deployed and validated services with Docker; used AWS CloudWatch for observability across logs and metrics, monitoring application health and triaging production and pre-production issues.
  • Wrote Python scripts, schedulers and automated workflows to handle recurring backend and operational tasks without manual intervention.
  • Expanded automated regression coverage using Jest and Cucumber, raising confidence in release cycles.
  • Covered frontend delivery in React and Redux for six months during a team resource shortfall, building and shipping UI features alongside backend work.
  • Mentored junior engineers through code review and backend design sessions.
Node.js · NestJSKafkaRedisDocker · AWSJest · Cucumber
Mar 2020 – Jun 2023
India

Career break: full-time preparation for the Indian Civil Services (UPSC) examination.

Jul 2018 – Feb 2020
Karlsruhe, Germany

Automotive Test Engineer

Robert Bosch GmbH (via Hays)

Validated automotive RADAR systems in C++ with the Vector toolchain in a safety-critical environment, working to formal test specifications and defect-tracking discipline.

C++Vector toolchain
Jun 2016 – Jun 2017
Frankfurt am Main, Germany

Test Engineer (Freelance)

Cint Consulting

Performed hardware-in-the-loop (HIL) testing of photovoltaic inverter systems in C++ using MATLAB/Simulink models and the Vector toolchain, validating controller behaviour against simulated grid and array conditions and automating regression test cases against formal test specifications.

C++MATLAB/Simulink
02
Selected projects
github.com/suryakapila
Retrieval · 01

RAG-Search-Engine

Built an information retrieval engine that incrementally evolves from BM25 to semantic, hybrid, and RRF retrieval, forming the foundation of a production-style RAG system. Explores the engineering trade-offs behind modern search systems rather than treating RAG as a black box.

Python · ChromaDB · BM25 · RRF
Agents · 02

Coding Agent

Designed and built a modular coding agent with separate planning, tool execution, and validation stages, enabling iterative code generation and verification before changes are applied.

Python · Gemini Flash · tool use
Pipelines · 03

AI Investment Pipeline

Designed and built an AI orchestration engine that coordinates three autonomous workflows for startup sourcing, thesis-based evaluation, and investment memo generation. Implemented structured outputs, caching, and run-level observability to deliver reliable, traceable, and repeatable results.

TypeScript · Claude API · ChromaDB
Messaging · 04

RabbitMQ Pub/Sub Service

Event-driven pub/sub service on RabbitMQ (TypeScript): topic and direct exchanges, dead-letter queues, publisher confirms, and prefetch-based backpressure, with three-way message acknowledgement for reliable delivery.

TypeScript · RabbitMQ
03
Skills
LanguagesTypeScript · JavaScript · Python · SQL · C#
Microsoft Stack.NET Framework · C# · WCF · ASMX · SQL Server · T-SQL · SOAP / XML web services · IIS
BackendNode.js · NestJS · Express.js · REST & SOAP API design · microservices · systems integration
FrontendReact · Remix
Distributed SystemsKafka · RabbitMQ · event-driven architecture · asynchronous processing
DatabasesSQL Server · PostgreSQL · MongoDB · Redis · TypeORM · Mongoose
Cloud & DevOpsAWS (EC2, S3, CloudWatch) · observability & monitoring · Docker · GitHub Actions · Jenkins · CI/CD
Testing & SecurityMSTest · Jest · Cucumber · unit & integration testing · OAuth2 · JWT · RBAC · Zod · secure coding
AI EngineeringClaude Code · Claude API / Anthropic SDK · Gemini Flash · RAG · ChromaDB · structured outputs · prompt engineering
ToolsGit · GitHub · Jira · Agile / Scrum
04
Resume MCP

Don't read the PDF. Query it.

This resume also ships as a Model Context Protocol server. Point Claude Desktop, Claude Code or your own agent at it and the whole work history becomes callable: filter roles by technology, search every line, get contact details, all as structured JSON instead of prose.

It's a small server, and that's the point. It's the same shape as the backend systems I build every day: a typed boundary, validated arguments, predictable payloads.

claude_desktop_config.json
{
  "mcpServers": {
    "resume": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://surya-resume-site.suryakapila.workers.dev/mcp"
      ]
    }
  }
}
Server sourceReadmeresume.json

Or connect over HTTP (no clone, no key)

hosted endpoint · Streamable HTTP
https://surya-resume-site.suryakapila.workers.dev/mcp

In Claude (Desktop or web): Settings, then Connectors, then Add custom connector, and paste the URL. In Claude Code: run the command above. Any Streamable HTTP MCP client works.

Tools exposed

get_summaryname, label, summary
get_sectionwork | education | skills | projects
search_by_skillskill

Resources

resume://fullapplication/json
Ask it in your own Claudeno key, no backend

Connect the MCP server above, then ask your own Claude anything you'd normally scan the resume for: depth on a technology, scale of a system, why the career break. Copy a prompt to get started.

Answers come from your model, grounded in this resume over MCP.3 tools · 1 resource
05
Education & contact
2012 – 2015
M.Sc. Electrical Engineering & Information Technology
Hochschule Darmstadt, Germany
2008 – 2012
B.Tech. Electrical & Electronics Engineering
Jawaharlal Nehru Technological University

Building products that need reliable backend engineering?

Emailsurya.kapila@gmail.comGitHubgithub.com/suryakapilaLinkedInsurya-prakash-rao-kResumeSurya_Kapila_Resume.pdf