RD

Rahul Dhiman

Backend · Search · AI

Available for select product and platform conversations

Rahul Dhiman

Software Engineer | Backend, Search, AI & ML Systems

I design high-throughput backend systems, AI-powered support platforms, and ML-driven forecasting products that make complex problems feel fast, reliable, and commercially useful.

Software engineer with ~4 years of experience building enterprise search, low-latency APIs, production-grade LLM workflows, and machine learning systems for predictive analytics and revenue intelligence. My work sits at the intersection of distributed systems, information retrieval, applied AI, and product execution.

GitHubLinkedInhi@rahuldhiman.comFounder-friendly, recruiter-friendly, execution-first

Current focus

Search, retrieval, and AI systems that hold up in production.

Rahul Dhiman

50K+

Production users

10M+

Documents indexed

91%

Forecast accuracy

65%

Latency reduction

Quick actions

Open the action palette to print, share, copy contact details, or send the resume without hunting through the page.

What I do

The resume story in one read.

A concise overview for hiring managers, founders, and teams that need both systems depth and product judgment.

I build the backend systems and machine learning pipelines that power search, support, and AI-driven customer experiences at scale. My strength is translating deep technical constraints — from low-latency APIs to predictive analytics models — into products that feel immediate and reliable for end users.

I am at my best on platform-heavy problems: search relevance, ingestion pipelines, API performance, time series forecasting, AI retrieval systems, fault tolerance, and agentic workflow automation that touches real business outcomes.

I am targeting senior software engineering and ML engineering roles where systems design, search, AI integration, predictive analytics, and execution quality all matter. The differentiator I bring is practical depth with a product mindset.

Scalable backend architecture & FastAPI services
Search relevance, indexing, and retrieval
Machine learning & time series forecasting
LLM systems with grounding, agentic orchestration, and LangGraph

Positioning

Technical depth with commercial intent.

The through-line across the work: systems that scale, interfaces that serve real teams, and AI that is measured by outcomes rather than novelty.

Enterprise search at scale

Shipped search and retrieval systems where performance, relevance, and uptime materially affect support efficiency.

AI that survives production

Focused on grounded LLM workflows with fallbacks, measurable outcomes, and operational confidence.

Predictive ML for revenue-critical decisions

Built time series forecasting and risk-scoring models that feed directly into pricing and overbooking decisions — where accuracy has a direct dollar impact.

Operator mindset

I care about latency, observability, failure modes, and maintainability as much as shipping speed.

Voice and multimodal systems

Built real-time voice pipelines that orchestrate speech, context, and response generation with tight latency budgets.

Experience

Built for scale, measured by impact.

A structured timeline of the roles, systems, and outcomes that define the current body of work.

01

Software Engineer

Grazitti Interactive (SearchUnify)

2023 - Present

Hybrid / Remote

Own backend and AI-heavy initiatives for an enterprise search platform used across customer support and knowledge workflows.

  • Designed and scaled enterprise search systems serving 50K+ users across large support environments.
  • Built indexing pipelines for 10M+ documents across CRM, knowledge base, and internal content systems.
  • Reduced p95 latency from 400ms to 140ms through query optimization, caching strategy, and API tuning.
  • Developed backend services in Node.js and Python handling 1K+ QPS under production workloads.
  • Led Salesforce Service Cloud integration that improved knowledge recommendations and cut ticket resolution time by ~25%.
  • Architected LLM-powered support automation using agentic workflows and RAG-based retrieval, reducing repetitive support queries by 30% and cutting average handle time by 25%.
  • Designed and deployed multi-agent AI orchestration (LangGraph) enabling complex, multi-step query resolution without human escalation.
  • Improved reliability with fallback logic, production monitoring, and alerting around search failure states.
Node.jsPythonFastAPISearchRAGLangGraphSalesforceAWS

02

Software Engineer

Grazitti Interactive

2022 - 2023

On-site / Hybrid

Worked on search infrastructure foundations, ingestion pipelines, and production debugging across distributed services.

  • Built REST APIs and ingestion pipelines for search-oriented backend systems.
  • Designed batch and streaming workflows for document ingestion across multiple data sources.
  • Reduced ingestion failures by ~40% with retry orchestration, validation, and stronger failure recovery.
  • Debugged production issues in distributed services and improved service resilience.
REST APIsIngestionDistributed SystemsValidationObservability

Projects

Case studies that prove taste, architecture, and execution.

Each project is framed around the problem, the system design decisions, and the business or operational impact.

Machine Learning & Predictive Analytics

Hotel-Based AI Forecasting & Revenue Intelligence System

Details

A full-stack hospitality intelligence platform combining time series forecasting, dynamic pricing recommendations, no-show risk scoring, and a conversational agentic AI layer for hotel management teams.

Problem solved

Hotel revenue managers lacked actionable, real-time intelligence to optimize occupancy, pricing, and overbooking decisions — relying on manual spreadsheets and intuition rather than data-driven forecasting.

Outcome

Achieved 91% occupancy forecast accuracy (MAPE <9%), reduced revenue leakage from no-shows by ~18% through ML-based overbooking optimization, and surfaced dynamic pricing recommendations that improved RevPAR by an estimated 12–15% in simulation. Delivered a natural language query interface that reduced time-to-insight for managers from hours to seconds.

Machine LearningTime Series ForecastingPredictive AnalyticsDynamic PricingRevenue ManagementAgentic AILangGraphHospitality TechMLOps
Discuss architecture

AI Systems

Voice-to-Voice AI Support Agent

Details

A low-latency conversational support pipeline that converts speech to action-ready answers in real time.

Problem solved

Support teams needed faster first-response automation without sacrificing conversational quality or operational guardrails.

Outcome

Delivered sub-2 second multi-turn interaction latency with session continuity, fallback handling, and voice-first orchestration.

Voice AILatencyAutomation
Discuss project

Product Automation

Text-Based AI Support for Salesforce Case Management

Details

A context-aware AI assistant embedded into case workflows to speed up responses and reduce manual effort.

Problem solved

Case agents needed better suggestions grounded in enterprise context, not generic text generation.

Outcome

Integrated AI suggestions into agent workflows and reduced case handling time by roughly 25 to 30 percent.

SalesforceSupportLLM
Request walkthrough

Retrieval & Ranking

Production RAG System

Details

A grounded question-answering system built over enterprise knowledge with careful ranking, chunking, and answer controls.

Problem solved

Generic LLM outputs were not reliable enough for enterprise support use cases where correctness and traceability mattered.

Outcome

Improved answer accuracy by ~35% through retrieval design, ranking strategy, chunk quality, and grounding discipline.

RAGEvaluationEnterprise AI
Talk architecture

Search Infrastructure

Multi-Source Enterprise Search

Details

A federated search experience spanning CRM, documentation, and internal knowledge systems.

Problem solved

Users were losing time switching across fragmented systems with inconsistent relevance and duplicate results.

Outcome

Designed ranking, deduplication, and multi-source indexing for 10M+ documents across enterprise environments.

Enterprise SearchSearch RelevanceScale
See capabilities

Expertise

Organized by capability, not buzzwords.

A signal-dense breakdown of the domains I ship in most often.

Systems and backend design

High-signal execution on APIs, throughput, reliability, and operational safeguards.

Search quality and data pipelines

Deep comfort with indexing, retrieval, relevance, and scale-sensitive search experiences.

Applied AI and machine learning

RAG systems, agentic flows, time series forecasting, and pragmatic ML integration into real user workflows.

Backend & Distributed Systems

Node.jsPythonFastAPIREST API DesignMicroservicesAsync ProcessingLow-latency Services

Search & Retrieval

ElasticsearchSolrIndexing PipelinesRankingQuery OptimizationDeduplication

AI & LLM Systems

RAG PipelinesAgentic WorkflowsLangGraphLangChainPrompt EngineeringMulti-Agent OrchestrationFallback LogicEvaluation & Evals

Machine Learning & MLOps

Time Series ForecastingPredictive AnalyticsXGBoostProphetScikit-learnFeature EngineeringModel EvaluationDynamic Pricing ModelsRevenue ManagementDocker-based Model Serving

Data & Persistence

PostgreSQLRedisSQLData ModelingChunking StrategiesKnowledge SystemsContext RetrievalTime Series Data

Cloud & Delivery

AWS EC2S3RDSLambdaCI/CDDockerProduction Monitoring

Product & Operations

Service CloudSupport AutomationCross-functional DeliveryReliability Engineering

Metrics

Numbers that make the story easy to scan.

The signal most recruiters and engineering leaders want to find quickly: scope, load, performance, and measurable change.

0+

Years building production systems

0K+

Enterprise users served

0M+

Documents indexed across systems

0%

p95 latency reduction on key flows

0%

Repetitive support queries reduced

0%

Occupancy forecast accuracy (MAPE <9%)

0%

No-show revenue leakage reduced via ML overbooking model

0%

Simulated RevPAR improvement from dynamic pricing recommendations

Education

Compact, clear, and in service of the work.

Formal education is presented with restraint so the professional signal stays in focus.

Completed

Masters of Computer Application

Kurukshetra University

Contact

Built to convert curiosity into conversation.

If you are hiring for backend platforms, search, AI systems, or product infrastructure, this site should make the next step obvious.

hi@rahuldhiman.com

Direct response path for opportunities, portfolio requests, and architecture conversations.

India · IST (UTC+5:30)

Open to full-time, consulting, and high-impact platform work

GitHubLinkedInShare profileSend resume