Guna Sekhar Dora Kovvuru

Software Engineer

Bengaluru, Karnataka, India8 yrs 6 mos experience
AI EnabledHighly Stable

Key Highlights

  • Led AI knowledge infrastructure for major credit unions.
  • Architected scalable systems reducing latency significantly.
  • Mentored teams and drove technical direction in fintech.
Stackforce AI infers this person is a Fintech and SaaS expert with strong capabilities in distributed systems and AI infrastructure.

Contact

Skills

Core Skills

Large Language Models (llm)Retrieval-augmented Generation (rag)Knowledge GraphsTechnical LeadershipSoftware ArchitectureDistributed SystemsMicroservicesFull Text SearchAmazon Web Services (aws)

Other Skills

TemporalNode.jsVertex AIVector DatabasesTypeScriptNeo4jGraphitiGoogle GeminiPrometheus.ioSpring BootCross-functional CollaborationsSystems DesignMongoDBScyllaDBSQS

About

Engineer with 8 years building distributed systems at scale, most recently leading production AI knowledge infrastructure and multi-agent LLM systems for voice and chat bots serving credit union members across North America. At Interface AI, I designed and owned the end-to-end AI knowledge platform powering assistants for 10+ production tenants and 50+ sales tenants: · Built the production entity-first RAG pipeline with a 5-LLM orchestration chain (routing → sentiment → entity retrieval → answer formation → channel-aware synthesis), validated at 30 parallel tenants handling 1,000–1,500 documents each · Designed and built a next-generation graph-based system (Neo4j, Graphiti) - p99 read latency down from ~2.5s to ~800ms, per-query LLM calls from 5 to 2, and ingestion time from ~75 to ~40 minutes · Shipped a nano-ingestion pipeline that makes a new prospect's voice bot demo-ready in under 60 seconds, enabling the sales team to independently pitch live AI demos at industry expos (GAC 2026) and directly generate 35+ new client conversations · Built onboarding automation letting integration and sales engineers spin up new tenants independently, eliminating engineering bottlenecks after the first 3 production onboardings At Uni Cards (founding engineer, SDE2 → SDE3, 4+ years), I built core financial infrastructure from scratch across credit card, lending, and investment product lines: · High-throughput event service processing 15,000 RPM on SQS, MSK, and ScyllaDB, reused as a foundation across all business lines · Durable billing platform and gateway-agnostic payment infrastructure across co-brand programs with SBM, Yes Bank, and Bank of Baroda · RBAC framework powering CRM and internal developer tools across all co-brand programs · Designed Reverse Penny Drop, Transaction History, and SIM Binding for fraud detection, each built for reuse across every business line at Uni · Tech lead and scrum master coordinating 15 engineers across backend, frontend, and QA for the SBM co-brand card go-live; mentored a team of 7 At Unbxd, eliminated cascading search failures with a 5× more reliable SolrCloud infrastructure via ZooKeeper-based multi-cluster isolation, and migrated legacy services from EC2 to Kubernetes. BITS Pilani grad, B.E. Computer Science and M.Sc. Economics, 2018. Amazon internship alumni. Exploring Staff Engineer and L6 opportunities in AI infrastructure, knowledge systems, distributed systems, fin-tech systems, and platform engineering. Reach me at gunasekhar.dora3@gmail.com or +91 9441121564

Experience

8 yrs 6 mos
Total Experience
1 yr 5 mos
Average Tenure
1 yr 2 mos
Current Experience

Interface.ai

Software Development Engineer 4

Apr 2025Present · 1 yr 2 mos · Bengaluru · Remote

  • Interface AI - AI-first voice and chat platform for credit unions across North America.
  • · Designed and owned end-to-end the AI knowledge platform for 10+ production credit union tenants and 50+ sales tenants, covering ingestion, entity modeling, retrieval, and versioned APIs serving live member traffic
  • · Engineered the production entity-first RAG pipeline with a 5-LLM orchestration chain, routing, sentiment, entity retrieval, answer formation, and a channel-aware synthesizer enforcing grounding, PII suppression, and safety rules
  • · Architected a graph-based knowledge system (Neo4j, Graphiti) with SPO facts, conflict detection, and versioning - p99 latency down from ~2.5s to ~800ms and LLM calls per query from 5 to 2
  • · Drove the architectural direction of the graph system by resolving customer tickets and joining client calls across the first 3 production onboardings, surfacing root-cause knowledge gaps and co-designing solutions with the product team
  • · Built and load-tested async ingestion pipelines (Gemini 3, Temporal) across 20 parallel tenants at 1,000–1,500 docs each, reducing full-tenant sync from ~75 to ~40 minutes
  • · Shipped a nano-ingestion pipeline making a new prospect's voice bot demo-ready in under 60 seconds, enabling the sales team to pitch live demos at industry expos and generate 35+ new client conversations
  • · Built a Prometheus observability SDK for the legacy non-agentic bot system and executed a zero-downtime citation format migration to wiki-style, eliminating engineering dependencies for both legacy tenants and CU admins
Large Language Models (LLM)Retrieval-Augmented Generation (RAG)TemporalKnowledge GraphsNode.jsVertex AI+6

Uni cards

2 roles

Software Development Engineer 3

Promoted

Jul 2023Apr 2025 · 1 yr 9 mos

  • · Tech lead and scrum master for the 6-month SBM Co-brand card go-live - coordinated 15 engineers across backend, frontend, and QA; averted a critical first-cycle statement generation failure through immediate data fixes, recovering full statement delivery for all users within 6 hours of the deadline
  • · Drove HLD/LLD and architectural evolution across the co-brand platform suite - bill payment, stateless CRM, and card delivery - for all programs with SBM, Yes Bank, and Bank of Baroda
  • · Designed and implemented a Temporal-backed durable billing platform with pluggable statement parsers for multiple bank formats - critical infrastructure ensuring reliable monthly billing cycles across all co-brand programs
  • · Mentored a team of 7 engineers; owned technical direction, design sign-offs, and onboarding across the platform domain
  • · Streamlined customer support workflows by identifying systemic ticket patterns and driving long-term product fixes - reduced open ticket volume by 40% across co-brand programs
  • · Led development and tech-team coordination across all co-brand programs (SBM, Yes Bank, Bank of Baroda) — building and evolving multi-tenant platforms covering billing formats (APIs & files), PG integrations (Razorpay, BillDesk), CRM workflows, and card management; drove alignment with bank tech teams through the full SDLC as tech lead
Technical LeadershipTemporalSpring BootSoftware ArchitectureCross-functional Collaborations

Software Development Engineer 2

Dec 2020Jun 2023 · 2 yrs 6 mos

  • · As a founding engineer, designed and implemented the KYC platform, Transactional Risk Engine, Lender Report Generator, Notification, and Bill Payment systems for Uni's Pay 1/3 and Pay 1/2 BNPL products from scratch - scaling to 600K+ MAUs using MongoDB, ScyllaDB, SCXML, Spring Boot, SQS, and Drools
  • · Architected multi-tenancy across all owned microservices (HLD & LLD) to support Uni's P2P product launch, significantly reducing engineering effort for future product go-lives; mentored 3 SDE1s through the full implementation
  • · Implemented compliance-driven changes across payments and lender reporting to align with RBI's Digital Lending Guidelines 2022 - zero-downtime rollout across all external partner integrations
  • · Co-authored critical product features end-to-end - Reverse Penny Drop, Transaction History, and SIM Binding for fraud detection - designed for full reuse across Lending, Co-brand CC, Personal Loans, and Investments
  • · Built a high-throughput event service averaging 15,000 RPM with 99.99% uptime and at-least-once delivery guarantee using SQS, S3, Lambda, ScyllaDB, and MSK - adopted as the core reliability layer across all Uni business lines
  • · Designed and built a scalable RBAC framework gating CRM access and developer tooling for 45 engineers across all 3 co-brand programs; automated reconciliation of 6 daily bank files (~50K rows/day) using Redshift, S3, FilePulse, and Kafka - adopted as the standard integration template across all bank partnerships
Distributed SystemsSystems DesignSpring BootAmazon Web Services (AWS)Microservices

Unbxd inc

Software Engineer

Jan 2019Nov 2020 · 1 yr 10 mos · Bengaluru, Karnataka, India

  • · Designed and implemented a multi-cluster SolrCloud infrastructure on Apache ZooKeeper to isolate collections and prevent cascading failures across the search platform - improved reliability by 5×
  • · Developed a stateless payload enricher service for the search microservices stack, enabling clean separation of enrichment logic from core search processing
  • · Optimized AWS infrastructure costs by dockerizing legacy microservices from bare metal EC2 deployments; developed Helm charts that significantly reduced deployment rollbacks in the QA environment
  • · Enhanced autosuggest precision by refining the analysis chain and implementing document-wise validation; improved search latency via Redis caching; integrated Datadog metrics into new microservices; led resolution of all customer-reported indexing pipeline incidents
Distributed SystemsKubernetesJavaApache SolrRedis

Justdial

R&D Software Engineer

Jun 2018Dec 2018 · 6 mos · Bengaluru, Karnataka, India

  • · Developed and optimized Elasticsearch search queries for Justdial's main AutoSuggest system using PHP SDKs - improving result relevance and query performance across the full business catalog
  • · Resolved 30+ critical ordering bugs in Justdial's legacy search engine, aligning results with business rules and improving overall search accuracy
  • · Developed shell and Node.js scripts to periodically bulk-dump data from SQL to MongoDB for ongoing catalog quality validation
  • · Designed and implemented the JDBooks AutoSuggest system end-to-end - data ingestion into MongoDB, Elasticsearch indexing, and search query development for efficient lookups
ElasticSearchMongoDBNode.jsPHPFull Text Search

Amazon

Software Development Engineer Intern

Jul 2017Dec 2017 · 5 mos · Bangalore

  • · Developed an automated weekly reporting service for tier-2/3 vendor business analysis within the Vendor Success Program at Amazon Retail Business Services, using DynamoDB, S3, and SES
  • · Designed and implemented RESTful APIs for a prioritization rule engine, with end-to-end test coverage using JUnit, Mockito, and TestNG
  • · Built a spoofing functionality for enhanced troubleshooting and debugging of vendor workflows
Amazon Web Services (AWS)JavaREST APIsDynamoDBJUnit

Birla institute of technology and science, pilani

Teaching Assistant

Aug 2016Dec 2016 · 4 mos · Hyderabad Area, India

Mormugao port trust

Intern

May 2015Jul 2015 · 2 mos · Headland, Sada, Vasco Da Gama, South Goa, Goa, India

  • Developed a file management system to track document movement across departments. Utilised a legacy SAP ABAP Dev Workbench for the same

Education

Birla Institute of Technology and Science, Pilani

B.E. (Hons.) — Computer Science

Jan 2013Jan 2018

Birla Institute of Technology and Science, Pilani

M.Sc.(Hons.) — Economics

Jan 2013Jan 2018

Sri Chaitanya Junior College, Visakhapatnam

High School

Jan 2011Jan 2013

Bhashyam School, Srikakulam

High School

Jan 2006Jan 2011

Stackforce found 100+ more professionals with Large Language Models (llm) & Retrieval-augmented Generation (rag)

Explore similar profiles based on matching skills and experience