K. Sricharan

Data Scientist | Generative AI Engineer

I'm K. Sricharan, a Data Scientist at Deloitte with 5 years of experience building Generative AI applications, RAG pipelines, multi-agent systems, MCP integrations, and intelligent automation. My work focuses on turning complex business requirements into reliable, measurable AI solutions.

Discuss a project ↗ Connect on LinkedIn

Hyderabad, India · Open to remote collaboration with US-based teams · Email: sricharan.karingula@gmail.com

5 yrs

Experience stated in resume

AI initiatives highlighted

93%

DeepEval score reported for FinPilot AI

80%

Ingestion latency reduction reported

Selected work

AI projects & case studies

Selected AI initiatives I have worked on, with project scope, technical implementation, and reported outcomes.

Documented experienceMulti-agent · Finance

FinPilot AI — Multi-Agent Financial Analysis Platform

A financial analysis workflow combining a supervisor agent, intent-based routing, multi-format ingestion, semantic retrieval, and deterministic calculations.

LangGraph ReAct orchestration with four intent categories.
FastMCP, SQLite, and local file drivers for three input formats.
FAISS semantic search combined with deterministic Python tax logic.
Resume-reported results: ~80% lower ingestion latency and 93% DeepEval score.

Potential business use: finance operations, statement analysis, spend intelligence, and internal knowledge assistants.

Documented experienceRAG · Enterprise search

ALAN-GPT — Multi-Document RAG Search Engine

A retrieval system designed to make technical and API documentation searchable across a large catalog of production machine-learning models.

Documentation indexing and vector search across 500+ production ML models.
Technical and API-focused retrieval for developer workflows.
Resume-reported results: 75% lower developer lookup time and 91% retrieval evaluation score.

Potential business use: searchable SOPs, product documentation, support knowledge bases, and engineering portals.

Documented experienceAudio AI · Automation

Automated Call-Script & Report Generator

A Generative AI pipeline that transforms recorded calls into structured scripts, summaries, and extracted financial details.

Processed 1,000+ audio call recordings.
Generated structured call scripts and summaries for relationship-management workflows.
Streamlined financial detail extraction and report generation.
Resume-reported reduction of 80% in manual documentation effort.

Potential business use: customer-call intelligence, sales operations, compliance documentation, and meeting summaries.

What I can help build

Services for growing teams

RAG & enterprise knowledge search

Document ingestion, chunking, embeddings, hybrid search, reranking, citations, and evaluation.

AI agents & workflow automation

Tool-using agents, LangGraph orchestration, MCP integrations, human approvals, and traceability.

Document & audio intelligence

Extraction, classification, summarization, structured outputs, and operational reporting.

LLM evaluation & reliability

Ragas/DeepEval evaluation, regression tests, guardrails, observability, and quality improvement.

Technical toolkit

Tools I work with

Programming & APIs

Python, SQL, FastAPI, REST APIs, Pydantic

Generative AI & agents

LLMs, prompt engineering, LangGraph, tool/function calling, MCP, Hugging Face

Retrieval & evaluation

RAG, embeddings, semantic search, FAISS, vector search, Ragas, DeepEval, guardrails

ML & delivery

PyTorch, scikit-learn, Pandas, NumPy, Streamlit, Git, Docker

Recognition

Awards & certifications

Two-time Deloitte Impact Award winner for process automation and BI innovation.
Client recognition for high-performance dashboards supporting strategic decisions.
Anthropic learning credentials include Introduction to Model Context Protocol and Claude Code in Action, plus four additional certifications (as listed in resume).

Education

Statistics & Computer Science

Bachelor of Science · Jagruti Degree and PG College

July 2018 – July 2021 · CGPA 8.9

How I approach engagements

From business problem to dependable AI workflow

01 · Discovery

Understand users, workflow pain points, data sources, constraints, and success metrics.

02 · Prototype

Build a focused proof of concept with clear assumptions and a representative evaluation set.

03 · Validate

Test retrieval, factuality, latency, cost, access controls, and failure paths.

04 · Deliver

Package the application, document deployment, and hand over maintainable code and operating guidance.

Have an AI workflow in mind?

Let's turn it into a working solution.

Share the business problem, current process, and what a successful outcome looks like. I can help scope a practical first milestone.

Email about a project LinkedIn profile

About TechStudio

TechStudio — AI engineering services

TechStudio is the company through which I provide practical AI engineering and software development services. We help founders, startups, and businesses turn ideas and operational challenges into working solutions through custom AI project development, RAG applications, AI agents, workflow automation, MCP integrations, document intelligence, and LLM evaluation.

Engagements can cover discovery, proof-of-concept development, integration with existing systems, testing, deployment support, and technical handover.

Contact TechStudio Connect with Sricharan