SWAPNIL KHARCHE/PUNE · IN/OPEN TO WORK

Hey, I'm Swapnil,a Problem Solver!

Software Development Manager and Technical Lead with 12+ years modernizing mission-critical platforms across Government, Healthcare, E-commerce, and Financial domains.

I lead multi-stream teams that ship secure, scalable systems consolidating legacy estates, building integration backbones, and embedding AI-driven engineering into delivery workflows. Currently at Deloitte Consulting.

Pune, India12+ years in industryDeloitte Consulting LLPM.S. CS · Florida State UniversityOpen to senior roles
12+
Years engineering & leadership
5K+
Users on modernized government platform
52
Partner interfaces unified in one hub
300+
Screens secured with audit & access governance

01About

/ engineering philosophy

I translate complex domain and architectural challenges into secure, scalable platforms that reduce risk, optimize cost, and accelerate time-to-value. The work I'm proudest of is rarely the new, greenfield kind — it's the careful modernization of systems people already depend on, done without breaking their day.

For 12+ years I've led delivery on platforms that touch real users at scale: a government modernization program serving 5,000+ users and millions of transactions; an integration backbone wiring 52 partner systems; AI-driven internal tooling that reclaims engineering hours every week. My team ships with quality because the practices, CI/CD discipline, structured documentation, audit-grade traceability are non-negotiable, not afterthoughts.

02Experience

/ 2014 — present

A decade of delivery work across government modernization, distributed systems, cloud-native engineering, and AI-driven tooling — with the leadership accountability to back it.

Software Development Manager
Deloitte Consulting LLP · USA
Jul 2016 — Present

Leading multi-stream modernization of mission-critical government platforms.

Direct accountability across engineering, architecture, AI integration, cloud, DevOps, and delivery process for a long-running government modernization program. Played a key role in securing a 14-year enterprise contract through architectural POCs delivered in client presentations.

Engineering Leadership & People
  • Led and mentored 4 cross-functional teams (15+ engineers, QA, designers) driving high delivery velocity while preserving quality and stakeholder confidence.
  • Established Jira UAT training and developer enablement initiatives — onboarding time down 30%, cross-team productivity up 15%, solution quality standards elevated across teams.
  • Directed stakeholder discussions, defect triage, and UAT governance — accelerating testing cycles and improving execution efficiency.
Architecture & System Design
  • Architected a 360° Unified Portal consolidating data from 40+ legacy screens — case access time down 70%.
  • Designed a centralized Interface Hub integrating 52 partner interfaces as the unified inbound/outbound integration layer.
  • Built audit trail and access governance components across 300+ screens — ensuring regulatory compliance, RBAC, and full traceability on a mission-critical platform.
Backend & Distributed Systems
  • Delivered REST APIs that cut data transfer latency by 40% and streamlined interdepartmental workflows.
  • Containerized Java applications with Docker; environment-related defects down 30%.
  • Designed and published a GraphQL API POC on AWS, influencing firm-wide adoption of modern API architectures.
  • Enforced quality via JUnit-based TDD, Log4j structured logging, and JMeter performance validation.
Cloud & Cloud-Native Engineering
  • Deployed cloud-native services on AWS EC2 leveraging S3, IAM, SES, CloudWatch — improved scalability and operational monitoring.
  • Containerized DB2 and SQL Server with Docker — accelerating environment setup and improving consistency across teams.
  • Eliminated 6–12 hours of manual documentation per screen with an AI pipeline auto-generating Excel storyboards from Figma and legacy UI metadata — 40–70× productivity gain.
DevOps, CI/CD & Automation
  • Automated 10+ complex batch processes — manual effort down 95%, saving 3 days per execution cycle.
  • Owned, maintained, and scaled CI/CD on Jenkins and AWS CodePipeline across multiple projects — eliminating the need for a dedicated DevOps engineer.
  • Automated database backup & archival workflows — storage and operational costs down 75%.
Scale, Data & Performance
  • Modernized a government platform serving 5,000+ users and millions of transactions — with full legacy parity and business continuity.
  • Ensured 100% availability of mission-critical batch jobs through proactive monitoring and issue resolution.
  • Built a checksum-based reconciliation tool validating DB2→Oracle migration across 300+ tables and millions of records — manual verification effort down 90%.
  • Migrated DB2 schemas, data, and queries to SQL Server using SSMA, fully automated, ensuring integrity end-to-end.
Research Assistant
Florida State University · Dept. of Economics
Oct 2014 — May 2016

Built a large-scale search engine over 100M+ scraped records.

Scraped and processed 100+ million web records in Python; built a standalone search engine that cut data access time by 96% and enabled faster research insights for the department.

Open-source contribution · jEdit
  • Implemented word-wrap functionality in Java to improve text editor usability.
  • Enhanced hyper-search results by highlighting lines by type — improving developer productivity.
  • Added a bug submission feature, streamlining issue reporting within the platform.

03Selected work

/ shipped & in production

A handful of recent projects that show how I think about scale, integration, and the responsible introduction of AI into engineering workflows. Built across the Deloitte modernization program.

01 / ArchitectureGovernment

360° Unified Portal: one screen for case workers, forty behind it.

Consolidated case-worker data scattered across 40+ legacy screens into a single unified portal. Designed the data aggregation layer, the security model, and the navigation paradigm so that case workers stopped tab-hopping and started doing the work.

JavaSpring BootRESTSQL ServerAWS
70%faster case access
40+screens consolidated
02 / IntegrationPlatform

Interface Hub — fifty-two partner systems, one integration layer.

Designed and shipped a centralized integration platform that absorbed all inbound and outbound enterprise data flows. Replaced point-to-point sprawl with a contract-first hub: easier to monitor, easier to govern, easier to extend.

JavaMicroservicesRESTKafkaSwagger
52interfaces integrated
1platform to govern them
03 / AI & EngineeringInternal Tool

RAG Documentation Assistant — semantic recall over OCR'd PDFs.

Built an internal documentation assistant using Flask and ChromaDB with semantic retrieval over OCR-processed PDFs. Made the team's institutional knowledge actually queryable — the kind of fix that compounds quietly across a quarter.

PythonFlaskChromaDBRAGOCR
75%+time-to-info reduction
2–3hreclaimed / dev / week
04 / AI OperationsProduction

Ticket Intelligence System — human-in-the-loop triage for 3K+ tickets/mo.

An AI-driven classification and routing system for support tickets, built on OpenAI, AWS Lambda, and n8n with a human-in-the-loop checkpoint. Routes 3,000+ tickets a month — routing time collapsed from hours to minutes.

OpenAIAWS Lambdan8nPython
91%routing accuracy
3K+tickets / month

04AI Engineering

/ pragmatic, in production

I treat AI as another tool in the engineering stack — useful when the problem fits, measured by the hours and dollars it reclaims, and instrumented like any other production system. Below are three patterns I've shipped.

PROJECT 01

Live RAG demo: Ask questions to any uploaded PDF, on this site.

A working Retrieval-Augmented Generation app right here: upload a PDF and get answers grounded in its contents, with citations and follow-up memory. Local parsing, embeddings, and reranking; streamed answers. Click to try it.

Next.jspgvectorTransformers.jsGroqOCR
Try it live →
PROJECT 02

Human-in-the-loop classification at production scale.

LLM-based triage with a confidence threshold and a human reviewer for the long tail. Designed for the 91%-accurate case to land automatically and the 9% to land safely.

OpenAILambdan8nPython
PROJECT 03

AI-driven artifact generation in delivery workflows.

Pipelines that auto-generate Excel storyboards from Figma and legacy UI metadata — eliminating 6–12 hours of manual documentation per screen and delivering 40–70× gains.

PythonFigma APIn8nExcel

05Capabilities

/ stack & practice

The tools, languages, and disciplines I reach for. Depth in Java/Spring and AWS; breadth across data, AI, DevOps, and the practice of running engineering teams.

Languages & Frameworks
JavaSpring BootMicroservicesPythonFlaskRESTGraphQLJPAHibernateAngularReactHTMLCSS
Cloud & Infrastructure
AWS EC2S3IAMLambdaSESCloudFrontCloudWatchCodeCommitAzure API ManagementAzure DevOpsMS Entra (SSO)DockerKubernetesTomcatWebSphere
DevOps & Quality
JenkinsAWS CodePipelineSonarQubeJaCoCoJUnitTDDJMeterMavenGitSVNTFSOpCon Scheduler
Data & Messaging
OracleSQL ServerMySQLDB2NoSQLKafkaChromaDB
AI & Intelligent Systems
RAGOpenAIGroq AISupabaseLLM IntegrationPrompt Engineeringn8nCUDAVector DBOCR Pipelines
Legacy & Mainframe
COBOLCoolGenCopyBooksJCLLog4j
Process & Leadership
Engineering ManagementMentorshipStakeholder MgmtAgile / ScrumSprint ManagementCompliance & Audit WorkflowsArchitecture ReviewsRoadmap Planning
Collaboration & Docs
JIRAX-RaySwaggerConfluenceFigmaDraw.ioExcelLovable

06Recognition

/ credentials & education

Certifications I've earned, awards I've been recognized with, and the academic foundation under all of it.

Certifications

Azure AI FundamentalsMicrosoft
Fundamentals of Accelerated Computing with CUDA PythonNVIDIA
Generative AI EngineerDeloitte
Generative AI Prompt EngineerDeloitte
Building Scalable Java Microservices with Spring Boot & CloudCoursera

Awards & Publications

Outstanding Performance Award — Coaching & Team LeadershipDeloitte
Applause Award — Automation impactDeloitte
White paper: Intranet Conferencing SolutionIJETAE

Education

M.S. in Computer ScienceFlorida State University · 2013–16
B.E. in Computer EngineeringPune University · 2008–12

07Contact

/ open to senior roles
Available for senior roles · Pune or remote

Building something non-trivial at the systems layer? I'd like to hear about it.

Open to Software Development Manager, Engineering Manager, Staff Engineer, Principal Engineer, and Technical Lead roles: full-time or contract. Pune or remote-friendly.