Software Engineer | Backend, Distributed Systems & AI

Mano Bharathi M

I build backend services, distributed systems, observability, and AI-agent workflows. I have 2+ years of experience at OpenText, including my internship.

2+ Yrs

Experience incl. Internship

Cloud Connection Capacity

75%

Fewer Escalations (Staged)

~60%

Less Manual Triage (Prototype)

View Resume Explore Projects

Core Technical Domains

Backend & Data

Python, Java, Go, C++, SQL, FastAPI, Spring Boot, PostgreSQL, Vertica, Redis, and Elasticsearch.

Distributed Systems & Infrastructure

REST APIs, WebSockets, Kafka, change data capture, Linux, Docker, Kubernetes, Helm, Git, and CI/CD.

AI Agents & Retrieval

LangGraph, FastMCP, MCP, RAG, embeddings, vector retrieval, tool calling, evaluation, and PyTorch.

Professional Experience

Associate Software Developer

OpenText · Bengaluru

Oct 2024 — Present

Building backend services, distributed cloud connectivity, observability, and AI-agent workflows at OpenText.

Case-Aware Dual-Memory Agent

Focus: Enterprise support using RAG and reusable case memory.

Built: A case-aware agent combining enterprise knowledge with memory of previous cases.

Evaluation: Staged evaluation of escalations, response latency, tokens, and LLM calls per ticket.

Escalations: Reduced by 75%, from 20 to 5.

Latency: Reduced by 43%, from 1,324 to 750 ms.

Efficiency: Tokens per ticket fell 33% (502 to 336); LLM calls fell 40% (1.0 to 0.6).

Cloud-Connectivity Capacity

Focus: Inter-region connections in a distributed cloud-connectivity service.

Context: The service previously supported 55 inter-region connections.

Approach: Redesigned UUID storage in the existing service.

Compatibility: Delivered without a database schema migration.

Capacity: Increased supported connections from 55 to 165.

Outcome: 3x cloud-connectivity capacity.

Time-Series Baseline Service

Focus: Scalable time-series queries for observability.

Built: A baseline service with dynamic upper and lower bounds.

Raw Data: 5-minute data for detailed queries.

Aggregates: Hourly aggregates for broader time ranges.

Routing: Adaptive selection between raw data and hourly aggregates.

Outcome: Supported scalable observability queries across data granularities.

CVE Automation Prototype

Focus: Reducing manual vulnerability triage effort.

Ingestion: Collected and normalized security advisories.

Matching: Matched advisory information against inventory.

Remediation: Connected the workflow to network automation.

Scope: An end-to-end prototype covering ingestion through remediation.

Outcome: Reduced manual triage effort by approximately 60%.

Software Engineering Intern

OpenText · Bengaluru

Apr 2024 — Sept 2024

Built Python and SQL automation for license-usage extraction and CSV reporting, replacing repetitive manual reporting workflows.

Selected Projects & Research

Capability Runner

Browser Automation

Built LLM-assisted workflow discovery, saved capability artifacts, deterministic replay, action validation, page-evidence checks, expected-error handling, and human takeover.

Python · Playwright · React · TypeScript · LLM Tool Calling

Project Details & Source

SyncStream

Distributed Systems

Built a CDC pipeline with Debezium, Kafka consumers for Redis and Elasticsearch, shared retries, dead-letter queues, and consumer-management APIs.

Java · PostgreSQL · Debezium · Kafka · Redis · Elasticsearch · Docker

Project Details & Source

Ticketless IT/HR Voice Support

Multimodal AI Agents

Built speech recognition, WebSocket voice interactions, VLM/OCR screen understanding, enterprise RAG, and LangGraph orchestration with human approval and structured verification before closure.

Speech Recognition · WebSockets · VLM/OCR · RAG · LangGraph

Project Details & Source

Adaptive Runbook Intelligence Platform

Agent Memory & RAG

Built knowledge retrieval, case memory, a runbook library, policy-based routing, feedback tracking, and a fast path for known-good runbooks without LLM calls.

Python · ChromaDB · SQLite · Streamlit · MCP · Embeddings

Project Details & Source

Adaptive Compute Efficient Learning via Conceptual-Criticality

AI Research

Co-authored the AAAI 2026 Student Abstract and contributed to separate notebook prototypes for criticality estimation and early-exit inference.

Python · PyTorch · Transformers · Jupyter · Early Exit

Project Details & Source

Dynamic vs. Fixed K Reranking in RAG

Retrieval & Evaluation

Built a notebook comparison using embeddings, FAISS retrieval, cross-encoder reranking, and fixed versus score-based dynamic context selection.

Python · Sentence Transformers · FAISS · Cross-Encoder · Jupyter

Project Details & Source

Awards & Recognition

CVE agent and customer demonstrations

12 February 2026

Put Customers First · OpenText

Recognized for developing a CVE data-extraction agent for a Policy AI assistant prototype and contributing to customer demonstrations.

Raw-data sleeve visualization

25 July 2025

Raise the Bar · OpenText

Recognized for delivering sleeve visualization for raw datasets on time in the first sprint of PI44.

Dark mode, reporting and Azure backend

14 February 2025

Raise the Bar · OpenText

Recognized for dark mode support, license-consumption reporting, and quickly becoming productive on Azure backend work.

AWS and Azure network visualization

24 January 2025

Raise the Bar · OpenText

Recognized for implementing AWS and Azure network-test visualization on the geographical map and demonstrating working code for early feedback.

View Awards in Resume