Sravanthi M

CEO

Hyderabad, Telangana, India17 yrs 4 mos experience

AI ML PractitionerHighly Stable

Key Highlights

17 years of experience in AI and automation.
Led architecture of scalable conversational platforms.
Mentored teams on LLM evaluation and AI strategy.

Stackforce AI infers this person is a SaaS-focused AI Engineer with expertise in conversational platforms and automation.

Contact

Skills

Core Skills

Conversational AiMicroservicesAutomationVoice Ai

Other Skills

AI & Automation InnovationsAI AgentsAI ApplicationsAlgorithmsData MiningData StructuresDatabase DesignElasticSearchFull-Stack DevelopmentGitGrailsHibernateInformation RetrievalJavaJavaScript

About

I’m a Principal AI Engineer with over 17 years of experience designing data-driven and intelligent automation systems—from early ML pipelines and knowledge extraction frameworks to today’s LLM- and agentic AI–powered conversational platforms. My journey has spanned building scalable data applications that extract actionable intelligence, architecting real-time voice and chat assistants, and designing reliable AI orchestration frameworks that combine large language models with deterministic workflow control. At TrueLark, I’ve led the architecture of real-time voice and chat assistants that unite GPT reasoning, retrieval pipelines, and event-driven microservices delivering measurable CX automation outcomes with 92% workflow accuracy and 90%+ F1-scores. My passion lies in designing production-grade AI orchestration frameworks balancing generative fluidity with deterministic control. I specialize in: - Multi-agent orchestration and real-time state management - Voice AI latency optimization (STT–LLM–TTS loop) - RAG pipelines, semantic retrieval, and guardrail design - Scalable microservices for high-volume AI workflows I also mentor teams on LLM evaluation, prompt engineering, and AI-driven product strategy, bridging research and production.

Experience

Truelark

2 roles

Principal Engineer

Promoted

Feb 2024 – Present · 2 yrs 1 mo · Hyderabad, Telangana, India · Remote

At TrueLark, I drive AI innovation in customer interaction automation, leveraging cutting-edge LLMs and microservices to deliver intelligent, scalable solutions.
🔹 AI & Automation Innovations
Architected UQRB Model (patent-pending), a groundbreaking automation system unifying FAQ handling, quick actions, and business notifications—achieving 95.68% precision and 93.65% recall.
Optimized workflow automation, increasing cancel automation by 41% and enhancing time detection accuracy by 58%.
🔹 LLM & Conversational AI
Architected and scaled a multi-agent conversational platform powering millions of chat and voice interactions.
Unified GPT-based reasoning, RAG pipelines, FSM control, and deterministic orchestration to improve automation reliability.
Built TrueLark’s first real-time voice assistant, integrating low-latency STT → LLM → TTS loops with caching and stateful orchestration.
Developed Chain-of-Thought filtering to improve response accuracy and reduce hallucinations.
🔹 Microservices & Scalable AI Architectures
Designed resilient microservices and real-time data pipelines enabling consistent, low-latency performance.
Established standards for prompt governance, model evaluation, and orchestration control across AI workflows.
Mentored cross-functional teams (ML, backend, product) to accelerate AI adoption and reliability.
Driving AI-powered customer support that feels natural, efficient, and scalable. Always exploring what’s next in AI-driven automation.
#AI #LLM #Microservices #Automation #ConversationalAI #VoiceAI #RAG #AgenticAI

Multi-agent SystemsRetrieval-Augmented Generation (RAG)Large Language Models (LLM)Voice BotsConversational AIAI Applications+1

Senior Software Engineer

Mar 2022 – Present · 4 yrs · Hyderabad, Telangana, India · Remote

Bank of america

Lead Analyst and Individual Contributor

Aug 2018 – Mar 2022 · 3 yrs 7 mos · Hyderabad, Telangana, India

We built a real-time monitoring service that predicted root causes for disputes that arise out of settlements between counter parties. Framework was generic and flexible enough that saved 60 to 90% of development time and around 60-80% of analysis time for analysts, thus saving millions of dollars in cost for around 9-10 projects.

Strand life sciences

4 roles

Senior Software Developer - II

Promoted

Apr 2012 – Aug 2018 · 6 yrs 4 mos

As a tech lead worked on developing an open source platform for cataloging India Biodiversity Information to showcase the richness of Indian Western Ghats biodiversity.
As part of it we built a citizen science crowd sourcing platform which collect biodiversity observation photographs from general public from all over India.
This open source project has been replicated as national biodiversity platform in Bhutan and Africa.

Senior Software Developer - I

Apr 2011 – Mar 2012 · 11 mos

Worked on several projects doing indexing, browsing and searching microscopy images data.
Tried to build a RDF based data store for company resources as hobby project.

Software Developer III

Promoted

Apr 2010 – Mar 2011 · 11 mos

Worked on building a competitive intelligence tool called Textminer to analyze various company profiles and extract facts about them. This product is an enterprise solution to mine client’s data store and extract relations between various entities. It also provides an intuitive UI for searching and visualizing these facts.

Software Associate

Aug 2008 – Mar 2010 · 1 yr 7 mos

Worked on a business intelligence tool to perform sentiment analysis on user reviews posted on web
forums for various market products. The sentiments extracted across various facets for these products
are visualized in a dashboard to help in making decisions.