AI ENGINEER / GENAI / ML SYSTEMS

Intelligence,
engineered for
the real world.

I design and build production AI systems across agentic workflows, RAG, Voice AI, and intelligent automation.

Explore my work Resume
BASED IN INDIABUILDING BEYOND THE DEMO
THE HUMAN BEHIND THE SYSTEMS
KS / 01ks.
Kevin SanganiAI Engineer · Ahmedabad
[ PROFILE IMAGE — TO BE ADDED ]
THE EMERALD OBSERVERAlways processing.
⌘   Agentic systems
Python at the core
SCROLL TO EXPLOREHUMAN CURIOSITY. ENGINEERED INTELLIGENCE.
01 / SELECTED AI WORK

Systems built for
real-world problems.

Selected client systems designed around real business workflows, intelligence, and production reliability.

PROJECT INDEX04 / SELECTED SYSTEMS
/ 01Voice & conversational AI

AI voice. Real action.

Conversations that don’t just respond — they take action.

An inbound and outbound AI voice platform connecting lead conversations, follow-ups, appointment scheduling, and tool execution into one continuous workflow.

PythonPipecatTwilioGeminiDeepgram
Explore case study Professional work at Wishtree Technologies
CONCEPTUAL SEQUENCELive system loop
CONVERSATION ENGINECONCEPT / 01
Demo schedulingInbound conversation
Live call
CALLER

I’d like to schedule a demo.

AI AGENT

Sure. Let me check availability.

TOOL EXECUTIONcalendar.check()
Available slot found Ready to schedule
ILLUSTRATIVE WORKFLOWPYTHON + AI SYSTEMS
  1. 01Listen
  2. 02Understand
  3. 03Execute
  4. 04Confirm
/ 02RAG & knowledge intelligence

Knowledge, connected.

Turning scattered knowledge into answers people can actually use.

A knowledge intelligence platform that transforms internal information into searchable context and grounded answers through retrieval-driven AI workflows.

PythonFastAPILangGraphQdrantPostgreSQL
Explore case study Professional work at Wishtree Technologies
CONCEPTUAL SEQUENCELive system loop
KNOWLEDGE ENGINECONCEPT / 02
PDFDOCXWEB
SEMANTIC RETRIEVAL
How does our onboarding work?
GROUNDED ANSWER

Relevant context. Connected knowledge.
An answer rooted in your sources.

Source 01Source 02
ILLUSTRATIVE WORKFLOWPYTHON + AI SYSTEMS
  1. 01Ingest
  2. 02Connect
  3. 03Retrieve
  4. 04Ground
/ 03Intelligent automation

From signal to action.

Complex workflows, turned into intelligent automation.

An AI-driven sales workflow connecting discovery, enrichment, contextual research, personalized content, human approval, and scheduled delivery.

PythonFastAPIGeminiApollo.ioApify
Explore case study Professional work at Wishtree Technologies
CONCEPTUAL SEQUENCELive system loop
WORKFLOW ENGINECONCEPT / 03
Outreach, orchestrated.From discovery to delivery
DiscoverConnected workflow
EnrichConnected workflow
ResearchConnected workflow
GenerateConnected workflow
ReviewHuman approval
ScheduleConnected workflow
DeliverWorkflow complete
Human judgment, in the loop.
ILLUSTRATIVE WORKFLOWPYTHON + AI SYSTEMS
  1. 01Discover
  2. 02Research
  3. 03Review
  4. 04Deliver
/ 04Multi-agent AI

One goal. Many perspectives.

Specialized agents working together, not just another prompt.

A multi-agent social media platform coordinating ideation, virality, SEO, community alignment, quality review, and platform-specific publishing.

Multi-Agent SystemsTool CallingOrchestrationStateful Workflows
Explore case study Professional work at Wishtree Technologies
CONCEPTUAL SEQUENCELive system loop
AGENT ORCHESTRATIONCONCEPT / 04
CONTENT
Viral agent
SEO agent
Community
Chief Curator
Specialized perspectives. Shared context.
Curated and ready for publishing
ILLUSTRATIVE WORKFLOWPYTHON + AI SYSTEMS
  1. 01Ideate
  2. 02Optimize
  3. 03Align
  4. 04Curate
02 / WHAT I BUILD

Intelligence,
put into production.

From a single interaction to an entire business workflow. Engineering the system around the intelligence.

01

Agentic AI

Agents with purpose, tools, and shared context.

Multi-agent systems / Tool calling / Stateful orchestration
02

RAG & Knowledge

The right knowledge, at the right moment.

Document ingestion / Semantic search / Grounded generation
03

Voice AI

Natural conversations. Meaningful outcomes.

Real-time voice / Telephony / Conversational workflows
04

Intelligent Automation

Complex processes, connected intelligently.

API integrations / Human review / Workflow execution
ALWAYS A STUDENT OF THE SYSTEM

Going deeper into
the intelligence.

Beyond building LLM applications, I continuously explore the foundations behind modern AI systems — from classical machine learning and neural networks to Transformer architectures, model training, fine-tuning, evaluation, and inference optimization.

MLDeep LearningAttentionTransformersLLMsFine-TuningInference
03 / AI ENGINEERING STACK

From model
to production.

A Python-first toolkit. Connected by systems thinking.
Explore a layer to see what powers it.

AI ENGINEERINGPython at the core.Models → systems → production
THE LLM SYSTEMS LAYER

LLM Systems

Language models connected to useful tools and real workflows.

OpenAIGeminiClaudeLlamaPrompt EngineeringTool CallingMCP
SUPPORTING ENGINEERING

TypeScript Next.js React Java Spring Boot PostgreSQL MongoDB REST APIs

04 / EXPERIENCE

Building AI
where it matters.

Real systems. Real workflows.
Engineering that meets the business.

DEC 2024 — PRESENTAHMEDABAD, INDIA

Wishtree Technologies

GenAI Engineer

Building production-oriented AI systems across conversations, knowledge, agents, and business automation.

  • Developed LLM agents, RAG pipelines, Voice AI systems, and intelligent workflow automations.
  • Built retrieval systems across document ingestion, embeddings, semantic search, caching, and response generation.
  • Engineered conversational AI and LLM infrastructure with Pipecat, LiveKit, Twilio, Ollama, vLLM, LangSmith, and Langfuse.
  • Connected AI systems to APIs, tools, databases, asynchronous workflows, and production business processes.
From model capability to business capability.
05 / MORE SYSTEMS

More things
I’ve brought to life.

Additional professional and client work at Wishtree Technologies. Different domains, the same drive to make AI useful.

FURTHER EXPLORATIONSExplore project focus
01

Multilingual Hospital Appointment Voice Agent

Voice AI

Multilingual conversations around hospital appointment workflows.

02

Healthcare AI Agents

Agentic AI

AI agents focused on healthcare workflows.

03

Personal Loan / Finance Agent

Conversational AI

Conversational assistance for personal finance workflows.

04

AI Avatar Support Agent

AI support

An avatar-based interface for AI assistance.

05

WishAI Internal Knowledge Assistant

Knowledge AI

An assistant focused on internal knowledge discovery.

06

Multiverz Multi-Agent System

Multi-agent AI

Specialized agents working within a shared system.

07

SalesBridge Automation Platform

Automation

Automation across connected sales workflows.

08

Shopify AI Automation SaaS

AI SaaS

AI-powered automation for commerce workflows.

09

HARDI Outreach Automation

Automation

Workflow automation focused on outreach.

Project focus shown. Detailed implementations remain private.

06 / ABOUT

Curiosity meets
engineering.

Kevin SanganiAI ENGINEER & CONTINUOUS LEARNER

I like solving problems where AI can remove complexity.

My interest in AI comes from solving real-world problems and transforming complex workflows into intelligent, automated systems. I enjoy working across the AI application lifecycle — from designing agents and retrieval systems to integrating tools, deploying infrastructure, evaluating outputs, and improving reliability.

At the same time, I’m going deeper into machine learning, deep learning, LLM architectures, model training, fine-tuning, and inference to better understand the intelligence beneath the applications I build.

01

Solve the real problem

Technology should reduce complexity, not simply demonstrate sophistication.

02

Build systems, not demos

An AI solution matters when it integrates reliably with real workflows and production environments.

03

Keep going deeper

Understand not only how to use models, but how they work, how they are evaluated, and how they can be improved.

THE FOUNDATION

Dharmsinh Desai University

Bachelor of Technology in Information Technology

2021 — 2025

CGPA 8.84 / 10