Available for opportunities
Ranjith Kumar K.N

Senior AI Engineer · Bengaluru, India · Open to Work

RANJITH

KUMAR K.N

Autonomous AI Systems · Multi-Agent Architectures · Enterprise LLMOps

0
Years Experience
0
Ops Reduction
0
Companies
View My WorkDownload Resume

01 · ABOUT

Who I Am.

I'm Ranjith Kumar K.N — a Senior AI Engineer specializing in autonomous multi-agent systems, frontier model orchestration, and enterprise MLOps pipelines that operate at production scale.

Currently at Access Healthcare, I architect enterprise-grade agent harnesses and procedural graph workflows (LangGraph) for healthcare operations and revenue cycles, alongside production GraphRAG architectures integrating Neo4j with hybrid vector retrieval.

Previously at Analytical Intelligence International, I developed autonomous multi-agent systems and self-evolving RAG pipelines. Earlier, I shipped petabyte-scale data lakes at Shodh AI, built real-time autonomous vehicle sensor fusion at Pravaig Dynamics, and delivered big data optimizations at Tata Consultancy Services.

Multi-Agent Architecture Expert
Enterprise MLOps Specialist
AIP Journal Published · Dog Emotion CNN
ranjith.config.ts

Education

MSc Big Data Analytics

St. Joseph's University · 2020–2022

BCA Computer Applications

Seshadripuram College · 2016–2019

Bengaluru, India

02 · EXPERIENCE

Where I've Built.

AH
Current Role

Senior AI Engineer

Access Healthcare•Bengaluru, Karnataka, India

July 2026 – Present
Architect enterprise-grade agent harnesses and procedural graph workflows (LangGraph, custom cyclic state machines) to automate complex, multi-stage healthcare operations and revenue cycle workflows, achieving 99.2% deterministic accuracy across dynamic clinical schemas.
Engineer production GraphRAG architectures integrating Neo4j knowledge graphs with hybrid vector retrieval (Weaviate, Pinecone), enabling multi-hop associative reasoning over unstructured EHRs and clinical notes with a 40% improvement in contextual precision.
Establish secure sandboxed execution environments and automated continuous evaluation harnesses (DSPy, Langfuse) to benchmark tool-calling fidelity, latency, and context retention across frontier reasoning models.
Direct model deployment pipelines utilizing vLLM, speculative decoding, and quantization (AWQ/FP8), slashing P99 inference latency by 45% while enforcing strict HIPAA compliance and zero-trust data boundaries through NeMo Guardrails.
LangGraphGraphRAGNeo4jWeaviatePineconeDSPyLangfusevLLMNeMo Guardrails
Impact ScoreHigh
AI

Senior AI Engineer

Analytical Intelligence International, LLC•Bengaluru, India

Jan 2025 – Jul 2026
40% reduction in manual processing time (multi-agent orchestration)
35% boost in retrieval precision (self-evolving RAG with Langfuse)
60% faster deployment cycles (zero-downtime MLOps pipelines)
99.9% system reliability at enterprise scale
GeminiGPT-4ClaudeAutoGenGADKMCPKubernetesTerraform
Impact ScoreHigh
SA

ML Data Engineer

Shodh AI Pvt Ltd•Bengaluru, India

Jul 2024 – Jan 2025
45% reduction in processing latency (petabyte-scale multimodal data lake)
70% faster deployment (MLOps with Kubeflow + Airflow)
40% increase in production reliability (drift detection + auto-retraining)
3x speedup in LLM fine-tuning throughput (Ray + PyTorch FSDP + LoRA)
Apache SparkDelta LakeLoRADeepSpeed ZeRO-3RayW&B
Impact ScoreHigh
PD

Data Engineer

Pravaig Dynamics Pvt Ltd•Bengaluru, India

Feb 2023 – Jul 2024
40% improvement in battery analytics accuracy (BMS/SOH estimation)
99.5% accuracy in autonomous vehicle object detection
Sub-100ms end-to-end latency across 500+ concurrent sensor streams
PyTorchKafkaMQTTROS2LiDARCAN-BUSKalman Filters
Impact ScoreMedium
VD

Data Engineer

Vumonic Data Labs Pvt Ltd•Bengaluru, India

Oct 2022 – Jan 2023
500GB daily telemetry processing at 99.9% uptime
35% processing efficiency improvement (dynamic Airflow DAGs)
20+ hours of weekly manual work eliminated
Azure DatabricksSnowflakePySparkAirflowS3
Impact ScoreMedium
TC

Big Data Engineer

Tata Consultancy Services•Bengaluru, India

Apr 2022 – Oct 2022
30% increase in Hadoop/Spark processing efficiency
45% reduction in ingestion errors (AWS Deequ + Great Expectations)
HadoopSpark AQEAWS DeequHDFS/YARN
Impact ScoreMedium

03 · SKILLS

What I Work With.

LLMs & Generative AI

Expert
GPT-4 / ChatGPT95%
Claude93%
Gemini92%
LLAMA / Mistral / Deepseek90%
RAG & Fine-tuning94%
PEFT / LoRA90%
Prompt Engineering95%
MCP / A2A88%

AI Frameworks & Orchestration

Expert
LangChain / LlamaIndex93%
AutoGen / CrewAI92%
HuggingFace / Transformers90%
Semantic Kernel85%
Vector DBs (Pinecone/Weaviate)88%
Stable Diffusion82%

ML/DL & Classical Modeling

Advanced
Python95%
PyTorch / Lightning88%
TensorFlow85%
Scikit-learn / XGBoost90%
OpenCV / NumPy / Pandas88%
TensorRT / Quantization80%
C++ / Java75%

Cloud & MLOps

Advanced
AWS (SageMaker, EC2, S3, EKS)85%
Azure (DevOps, AI Foundry, AKS)87%
Docker / Kubernetes90%
MLflow / Jenkins / CircleCI85%
Airflow / CI/CD88%
Terraform / Crossplane82%

Big Data & Data Engineering

Advanced
Apache Spark / Delta Lake90%
Kafka / MQTT88%
Hadoop / Hive82%
PostgreSQL / MongoDB87%
Redis / Elasticsearch85%
FastAPI / ETL Pipelines88%

Data & APIs

Proficient
SQL / NoSQL88%
ClickHouse / Aurora RDS82%
REST / GraphQL APIs85%
Real-time Processing80%
API Security / AWS WAF78%
LangGraphNeo4jvLLMLangChainAutoGenMCPGADKPineconeWeaviateLoRADeepSpeedRayW&BLangfuseKubeflowDelta LakePyTorch FSDPLangGraphNeo4jvLLMLangChainAutoGenMCPGADKPineconeWeaviateLoRADeepSpeedRayW&BLangfuseKubeflowDelta LakePyTorch FSDP
KubernetesTerraformSparkKafkaROS2Delta LakeDockerAzure DevOpsAWS SageMakerMLflowAirflowCI/CDGreat ExpectationsSnowflakeKubernetesTerraformSparkKafkaROS2Delta LakeDockerAzure DevOpsAWS SageMakerMLflowAirflowCI/CDGreat ExpectationsSnowflake

04 · PROJECTS

What I've Shipped.

Published Research

Dog Emotion Recognition Using Deep Learning

AIP PUBLISHED · 2022

Developed a specialized emotion classification system using three distinct CNN architectures and a proposed MBCC-CNN model. Achieved superior accuracy in emotional recognition tasks — published in the American Institute of Physics journal.

3

CNN Architectures

MBCC-CNN

Architecture

AIP Published

Status

CNNDeep LearningComputer VisionMBCC-CNN
View GitHub →Read Paper →
LLM · RETRIEVAL-AUGMENTED GENERATION

Math Solver with RAG

Modular mathematical problem solver implementing RAG with Pythia LLM. Context-aware retrieval for complex problem-solving.

RAGPythia LLMRetrievalMath Reasoning
View Project →
LLM · RAG · VECTOR DATABASES

LLM Connector Context

Universal search and agent integration engine that connects external data sources like Slack, Notion, and GitHub to vector databases and LLM contexts.

LLMRAGQdrantVespaVector DB
View Repository →
MULTI-AGENT · WEB EXTENSION

AI-Powered Web Automation

Multi-agent AI-powered Chrome, Firefox, and Edge extension that collaborates to plan, navigate, and validate complex browser tasks.

LLMWeb ExtensionReasoningBrowser Use
View Repository →
LLM · SQL ENGINE · AGENTIC DATA

Structured Query Engine

High-performance semantic engine and agentic data layer that translates LLM and natural-language queries into optimized SQL using Rust, Python, and MCP.

LLMSQL EngineReasoningExecutionMCP
View Repository →

05 · CONTACT

Let's Connect.

Get in Touch

Email

ranjith.kumar.kn26@gmail.com

Location

Bengaluru, India

LinkedIn

linkedin.com/in/ranjith-kumar-58a676191

GitHub

github.com/rangow4562

Available for opportunities
Usually responds within 24 hours

Currently accepting senior AI engineer and AI architect roles. Experienced with FAANG-style technical interviews and enterprise clients. References and code samples available on request.

Quick Actions

Email Me

ranjith.kumar.kn26@gmail.com

Schedule a Call

Book a 30-min intro

Download Resume

PDF · Updated 2026

LinkedIn

Connect professionally

Why Ranjith

5+

Years in AI/ML

6

Companies Shipped

99.9%

System Reliability

Let's Talk →

© 2026 Ranjith Kumar K.N. Built with precision.

Designed for recruiters who valueimpact
RK
AI ChatPowered by FreeLLM

Hello I am Ranjith

how can I help