▶ PORTFOLIO.EXE
LangGraphGraphRAGAgentic AIFastAPINeo4j

Hi, I am

AI Solutions Engineer

AI Engineer

Loksai_Palyam

ONLINE

Building agentic AI systems, GraphRAG pipelines, and production-grade backends for real-world impact.

loksaipalyam.dev@gmail.com+91-9701759158

India

REN
loksai@ai-solutions ~

$ whoami

Loksai Palyam — AI Solutions Engineer

$ cat contact.json

{
  "email": "loksaipalyam.dev@gmail.com",
  "phone": "+91-9701759158",
  "location": "India",
  "github": "https://github.com/loksai-dev"
}

$ ./spawn.sh --mode=explore

▮ Ready. Swipe → to explore →

> LS -L ./SKILLS

drwxr-xr-x 5 loksai staff 4096 tech_stack/

💻

Languages

PythonTypeScriptRustSQL
🧠

AI & Graph

Agentic AILangGraphLangChainLlamaIndexGraphRAGMCPHugging FaceNeo4jQdrant
⚡

Backend

FastAPIAsyncIORESTWebSocketsSQLAlchemygRPC
☁️

Cloud & DevOps

AWSDockerKubernetesTerraformGitHub Actions
📊

Eval & Data Ops

LangSmithRagasClickHouseAirflowdbtPrometheus

> CAT ./EXPERIENCE.LOG

5 entries — tap any role to expand

[PROFESSIONAL]

  • ▸Resolved 10+ critical incident tickets across AWS data pipelines, dbt transformations, and middleware integrations for the Air Liquide enterprise uDP migration initiative.
  • ▸Architected an end-to-end AI-powered Supplier AP AutoReply System using Google Gemini 2.5 Flash, Apps Script, and ServiceNow APIs — automating multilingual supplier inquiry classification, NER, and intelligent P2P ticket routing across APAC affiliates.
  • ▸Reduced manual AP ticket handling by 70% and saved 160+ operational hours/month for the GBS APAC P2P team through automated real-time payment status workflows.
  • ▸Engineered LLM prompt pipelines with context-aware classification logic supporting multilingual queries (Japanese, Mandarin, English) at production scale.

[LEADERSHIP & COMMUNITY]

> ./FUTURE --preview

Flagship build + roadmap ahead

FLAGSHIPPOC

Aegis Earth

AI-Native Subsurface Digital Twin Platform

An AI-native subsurface digital twin that unifies geotechnical reports, borehole logs, GIS data, utility records, and simulated sensor streams into one engineering intelligence system.

PythonFastAPILangGraphNeo4jQdrantPostgreSQL+5 more

→ Swipe to POC for full breakdown

Live IoT Integration

Connecting Aegis Earth to real-world sensor streams and live GIS feeds for continuous subsurface monitoring.

Multi-Agent Ops

Scaling LangGraph agent pipelines for autonomous ingestion, risk scoring, and engineering report generation.

Physics + AI Fusion

Advancing PINN models with DeepXDE to enforce geotechnical constraints in real-time predictions.

3D Twin Expansion

Enriching React Three Fiber visualizations with dynamic heatmaps, tunnel planning, and utility overlays.

> LS ./POC/

4 proof-of-concept builds — tap to expand

PythonGroq LLMLangGraphNeo4j AuraDBFAISSPM4PyFaster-WhisperCoqui TTSStreamlitFastAPIDocker
  • ▸Developed an enterprise-grade Agentic AI platform that combines Voice AI, Process Mining, Knowledge Graphs, GraphRAG, and Multi-Agent Reasoning to analyze business processes through natural language conversations.
  • ▸Built a hybrid retrieval architecture using Neo4j Knowledge Graphs and FAISS Vector Search for contextual insights from process execution data, SOPs, audit reports, and operational policies.
  • ▸Implemented a LangGraph-based multi-agent system with Intent Analysis, Process Mining, Knowledge Graph Retrieval, Document Intelligence, Root Cause Analysis, Recommendation, and Simulation agents.
  • ▸Generated and analyzed 50,000+ synthetic process cases using PM4Py for bottleneck detection, variant analysis, SLA monitoring, conformance checking, and throughput analysis.
  • ▸Designed a Voice AI pipeline using Faster-Whisper and Coqui TTS for real-time voice-driven process intelligence and conversational analytics.
  • ▸Developed interactive Streamlit dashboards for KPI monitoring, process exploration, root-cause analysis, knowledge graph visualization, and what-if process simulations.
PythonFastAPILangGraphNeo4jQdrantPostgreSQLDeepXDEPyTorchNext.jsReact Three FiberDocker
PythonFastAPILangGraphOllamaFaster-WhisperKokoro TTSQdrantPostgreSQLRedisNext.jsDocker
PythonNeo4j AuraDBNeo4j GDSGraphRAGLangGraphFastAPIKafkaDockerKubernetes

> GREP -r ./RESEARCH/

Papers, experiments & findings

> CERTIFICATIONS

Professional Certifications & Technical Achievements

ACHIEVEMENT ROOM — 1 UNLOCKED

EARNED BADGES

Neo4jGraph DatabaseKnowledge GraphCypherGraphAcademy

> OPEN ./GALLERY/

Visual artifacts & memories

> SIDE QUESTS

Optional adventures outside the main storyline

📸

Fashion Portrait Photography

Capturing style, light, and character in portrait frames

✒️

Calligraphy

Practicing letterforms with ink, rhythm, and precision

📚

Reading Books

Exploring ideas across tech, fiction, and philosophy

🎾

Tennis

Court sessions for focus, footwork, and endurance

SWIPE LEFT OR → TO EXPLORE