Senior Software Engineer @ Sabre — Available for high-scale backend challenges

Autonomous systems builder.Architecting for scale, concurrency, and mechanical sympathy.

Senior Software Engineer at Sabre, building high-throughput distributed travel infrastructure and next-generation commerce architectures. Proven track record of independently engineering mission-critical systems from zero—delivering sub-millisecond tail latencies, robust concurrency primitives, and zero-defect enterprise rollouts.

Bengaluru, India•IST (UTC+5:30)
01 // PROOF OF WORK

Flagship Systems & Infrastructure

Production-grade distributed tools architected from zero to eliminate bottlenecks in observability, API evolution, and latency.

[01]

Overwatch

Application-Level State Integrity & SLA Assertion Engine

Source Repository
PROBLEM STATEMENT

In modern event-driven architectures, infrastructure-level monitoring tools (Prometheus, Datadog) track container health, CPU utilization, and network throughput, but remain blind to application-level state drift. Overwatch bridges this observability gap by asserting state integrity at the data layer against hard operational SLAs.

Concurrency & Threading

Single background ticker driving an in-process Hierarchical Timing Wheel combined with bounded worker pools to eliminate Go scheduler timer-heap lock contention.

Protocols & Serialization

Consumes JSON/Avro CDC envelopes over Kafka partitions and delivers typed alerts via HTTP/2 server-side streaming gRPC using Protocol Buffers.

State & Memory Efficiency

Embedded, process-local PebbleDB LSM spool for zero-network-hop crash recovery paired with in-memory doubly-linked wheel buckets for O(1) event scheduling.

Benchmarks & Empirical Wins

Delivers O(1) timer operations with sub-millisecond local PebbleDB spooling while collapsing hundreds of point lookups into batched Elasticsearch _mget calls.

GoPebbleDBElasticsearchgRPCKafka
Scheduling Complexity:O(1) Constant Time
Persistence Hop:0 ms Local LSM Spool
Alert Streaming:gRPC HTTP/2 Server Stream
[02]

raqs (Response Agnostic Query System)

Dynamic Proxy for Zero-Disruption Backward Compatibility & Schema Evolution

Source Repository
PROBLEM STATEMENT

Continuous API evolution is non-negotiable in modern engineering, yet maintaining backward compatibility remains an expensive, labor-intensive barrier. raqs is a high-throughput dynamic proxy architecture designed to eliminate client-side disruption entirely by intercepting traffic and executing on-the-fly schema transformations—allowing legacy clients to query deprecated contracts while upstream microservices evolve freely.

Concurrency & Threading

M:N cooperative scheduling with bounded fan-out worker pools, context-aware propagation, and graceful lifecycle cancellation.

Protocols & Serialization

gRPC / HTTP/2 multiplexing with Protobuf v3 and zero-allocation code generation for transparent protocol passthrough.

State & Memory Efficiency

Zero/low-allocation buffer recycling via sync.Pool and non-blocking synchronization primitives to preserve cache locality.

Benchmarks & Empirical Wins

Sustained 75,000 – 100,000 RPS per node (8 vCPU / 16GB RAM). Latency: p50 < 1.2ms, p99 < 8.5ms under peak concurrency. < 1% CPU execution time spent in STW GC pauses.

GogRPCRedisProtobuf v3
Throughput / Node:75k – 100k RPS
p99 Latency:< 8.5 ms
STW GC Overhead:< 1% CPU
02 // AGENTIC AI & AUTONOMOUS SYSTEMS

Autonomous Agents & MCP Architecture

Engineering deterministic agentic pipelines, Model Context Protocol (MCP) server tooling, and semantic vector retrieval for distributed software systems.

FEATURED AGENT PLATFORM

Project ARES

Enterprise Autonomous AI Agent Platform for Architectural Governance & Compliance Verification

Bun • Go • pgvector • Ollama

An enterprise-grade autonomous AI agent platform that bridges live codebases and specification documents with LLM reasoning agents. ARES automates architectural governance, implementation planning, and pre-merge compliance verification across distributed software systems.

Multi-Service Agent Architecture

ares-backend
Orchestration & Planning Engine
Go (Golang) • Bounded Worker Pools • pgx / pgvector
ares-worker
ETL, Chunking & Embedding Pipeline
FastAPI • Python 3.12 • Git Connectors
ares-mcp-server
Model Context Protocol (MCP) Server
Bun • TypeScript • Stdio Transport
postgres-cluster
Vector Database & Entity Storage
PostgreSQL 16 • pgvector (768-dim)
inference-sidecar
Local Offline Inference
Ollama • nomic-embed-text • llama3

Autonomous Agent Capabilities

Context-Aware Architectural PlanningRAG & Planning

Dynamically retrieves local codebase structures and external documentation (Notion, Rally) using pgvector cosine similarity search to automatically draft architecturally compliant implementation blueprints.

pgvectorCosine SimilarityNotion API
Automated Verification & Compliance GuardrailsGuardrails

Reviews proposed code changes via git diff streams against reference architecture context and policy manuals, acting as an automated gateway before code merge.

Git Diff AnalysisAST Policy EngineAutomated Gateways
Hybrid Inference Routing (Cloud & Edge)Inference

Dynamically routes queries to the GitHub Copilot CLI for cloud LLM reasoning, or falls back to a fully offline local sidecar container running Ollama (nomic-embed-text & llama3).

GitHub Copilot CLIOllamaZero-Egress Fallback
Agentic Model Context Protocol (MCP)MCP Standard

Implements process-level stdio MCP server bindings, allowing developers to plug ARES governance and verification tools directly into modern agentic IDE workflows (Cursor, Windsurf, Claude Desktop).

BunTypeScriptStdio TransportJSON-RPC
Model Context Protocol (MCP) Server Tools
Bun Stdio Transport • Cursor / Claude Desktop Compatible
⚡ tool: ares_planning

Generates a detailed, architecturally compliant implementation plan based on requirements or Notion spec URLs.

Schema: { "taskDescription": "string (spec description or Notion URL)" }
⚡ tool: ares_verification

Analyzes a git diff stream against design standards and verifies architectural compliance before PR merge.

Schema: { "gitDiff": "string (unified diff stream)", "projectId": "UUID (optional)" }
ENTERPRISE AGENTIC PRODUCTION RECORD

At Sabre, pioneered the autonomous AI pipeline with deterministic orchestration, delivering zero-defect production deployments for IATA Offers and Orders.

View Enterprise Case Study
03 // PRODUCTION TESTED

Enterprise Scale & Impact

Engineering high-throughput infrastructure at Sabre, zero-defect production rollouts, and resilient distributed architectures.

Scale
5,000 req/s
Sustained Peak Cluster Load

Handled across distributed microservices during global search surges with linear scaling.

Optimization
83.3% p99 Drop
Latency Reduction

p99 tail latency plummeted from 30ms to 5ms with zero post-release production defects on customer cutover.

Security
97.35%
Bot Anomaly Detection

Real-time passive ML anomaly detection powered by Vertex AI evaluating TLS/HTTP/2 fingerprints.

Precision
0.1% FPR
False Positive Rate

Sub-50ms API SLAs maintained by dynamically throttling client Proof-of-Work friction exclusively for malicious probes.

Architecture Case Studies & Production Wins

Click to expand details
Bottleneck / Challenge:

Massive cache invalidation spikes caused thousands of concurrent threads to simultaneously query the underlying database (thundering herd), driving downstream connection-pool exhaustion and severe p99 latency degradation.

Engineered Solution:

Engineered distributed locking using Redis and Redisson with automatic watchdog renewals and lease timeouts. A deterministic single-leader thread repopulates the expired cache key while concurrent requests await lock release or serve stale-while-revalidate fallback data.

VERIFIED PRODUCTION OUTCOME

Downstream database query spikes during key expiration dropped to O(1). Stabilized p99 latencies under heavy load and eliminated cascading thread-pool starvation across the cluster.

#Redis#Redisson#Distributed Locking#Thundering Herd Mitigation#Stale-While-Revalidate
04 // THOUGHT LEADERSHIP

Publications & Systems Deep Dives

Peer-reviewed research in distributed systems, graph data structures, and continuous technical analysis.

Peer-Reviewed Academic Publications

2025 17th International Conference on Electronics, Computers and Artificial Intelligence (ECAI)•July 2025

Evaluation of Machine Learning Algorithms for Predicting Cybersecurity Incidents

View Paper

Evaluates supervised machine learning algorithms across the public GUIDE dataset to proactively forecast cybersecurity incidents and attack vectors, analyzing historical event signals to preempt large-scale operational system disruptions.

Machine LearningCybersecurityPredictive AnalyticsGUIDE DatasetAnomaly Detection
2022 4th International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT / IEEE Xplore)•December 2022

Ledger-based Electronic Voting System using a multi-tier security approach

View Paper

Proposes a tamper-evident, decentralized electronic voting architecture combining blockchain ledger immutability with graph database relationships to solve security, accessibility, and transparency bottlenecks at national election scale.

BlockchainGraph DatabaseNeo4jDistributed LedgerSecurity Engineering
Springer LNDECT (Lecture Notes on Data Engineering and Communications Technologies, Vol. 142)•September 2022

Patent Recommendation Engine Using Graph Database

View Paper

Enhances the depth and throughput of patent metadata extraction by modeling multi-dimensional citations as a graph, computing Jaccard distance metrics to uncover high-confidence semantic relationships while drastically slashing research manpower.

Neo4jGraph DatabaseJaccard DistanceData EngineeringRecommender Systems

Engineering Essays & Platform Outposts

05 // TECHNICAL ARSENAL

Battle-Tested Stack & Primitives

Core technologies and mechanical sympathy patterns utilized to engineer production systems.

Languages & Core Runtimes

Primary daily drivers for systems and high-throughput backend development.
Go (Golang)

High-concurrency systems, low-latency microservices, timing wheels

TypeScript

Type-safe distributed APIs, full-stack tools, Zod schemas

JavaScript (Node.js)

Event-loop architectures, streaming I/O, asynchronous services

Agentic AI & Autonomous Systems

Model Context Protocol (MCP) tooling, vector retrieval, and deterministic agent execution.
Model Context Protocol (MCP)

Stdio & SSE server tool bindings for Cursor & Claude Desktop

pgvector & Embeddings

Cosine similarity search, 768-dim high-dimensional indexing

Hybrid Inference Routing

Copilot CLI cloud routing with offline Ollama fallback sidecars

Deterministic Orchestration

State consistency, execution graphs, and compliance guardrails

Distributed Systems & Concurrency

Architectural patterns and communication backbones for resilient scale.
gRPC & HTTP/2

Multiplexed unary and server-side streaming RPCs

Kafka

Event streaming, CDC pipelines, partition concurrency

Redis

Distributed locking (Redisson), cache strategies, low-latency state

Hierarchical Timing Wheels

O(1) high-frequency timer and task scheduling

Bounded Worker Pools

Backpressure management and goroutine leak prevention

Distributed Locking

Watchdogs, lease timeouts, thundering herd mitigation

REST APIs

Hypermedia and contract-driven API engineering

Storage, Caching & Data Engines

Specialized persistence layers matched to workload mechanical sympathy.
PostgreSQL

ACID transactions, indexed relational schemas, connection pooling

MongoDB

Document persistence, aggregation pipelines, replica sets

PebbleDB (LSM)

Embedded process-local write-ahead LSM spool for instant crash recovery

Elasticsearch

Inverted index search, batched _mget retrieval, log analytics

Neo4j

Graph algorithms, citation topologies, traversal patterns

Cache Invalidation Patterns

Stale-while-revalidate, probabilistic early expiration, warming

Infrastructure, Reliability & Systems

Production-tested deployment, orchestration, and Unix runtime mastery.
Docker

Multi-stage minimal image packaging and container isolation

Linux Internals

Thread scheduling, memory allocation, network socket tuning

Git & GitLab CI/CD

Automated test validation, quality gates, zero-downtime rollouts

AWS Services

SES v3, S3, IAM, CloudFront, Route 53

Resilience Patterns

Bulkhead isolation, sliding-window circuit breakers, token bucket

Frameworks & Modern Web

Modern tools for API routing, validation, and full-stack interactive delivery.
Next.js (App Router)

React 19 Server Components, streaming SSR, API routes

Gin

High-performance Go HTTP microservices and middleware

Express

Lightweight Node.js routing and proxy interceptors

Tailwind CSS

Utility-first design systems and strict dark palettes

Zod

Strict runtime schema validation and contract enforcement

06 // TRAJECTORY & CREDENTIALS

High-Velocity Engineering Trajectory

Rapid vertical progression at Sabre, ongoing graduate systems research at UT Austin, and national competitive honors.

Sabre Career Ladder (3 yrs 10 mos)

Intern → Senior SE

Senior Software Engineer

Current Focus
Aug 2026 – Present•3 mos
Sabre • Bengaluru, India
  • ›Pioneered the team's Agentic AI framework, engineering an autonomous AI pipeline backed by a deterministic orchestration layer for reliable state management, execution guarantees, and production readiness.
  • ›Co-designed resilient, high-throughput SabreMosaic backend services alongside Principal Architects to power mission-critical IATA Offers and Orders commerce workflows.
  • ›Served as Technical Bridge across partner engineering pods and led a high-performing Scrum team of 5 engineers, conducting 50+ technical design and PR reviews.
  • ›Enforced automated validation pipelines and robust CI/CD quality gates, achieving client production rollout with zero post-release defects.

Software Engineer III

Oct 2025 – Jul 2026•10 mos
Sabre • Bengaluru, India
  • ›Architected next-generation travel technology services for Sabre Digital Experience, focusing on fault-tolerant, low-latency microservices.
  • ›Designed and implemented a complex seat map module alongside system architects, optimizing memory layout and upstream data flow.
  • ›Mentored junior engineers and guided 3 interns through their successful transition to Full-Time Engineers (FTEs).

Software Engineer II

Jul 2024 – Oct 2025•1 yr 4 mos
Sabre • Bengaluru, India
  • ›Engineered backend services for Sabre Digital Experience and modern airline commerce (IATA Offers and Orders).
  • ›Delivered high-resiliency seat map subsystems under stringent latency SLAs.
  • ›Authored internal developer guides and standard operating procedures to streamline team onboarding.

Software Engineer I

Jul 2023 – Jul 2024•1 yr 1 mo
Sabre • Bengaluru, India
  • ›Maintained and modernized Sabre Kiosk, the flagship enterprise solution for airport self-check-in services.
  • ›Built B2C booking and reservation management capabilities for the Digital Experience travel commerce platform.

Associate SDE Intern

Jan 2023 – Jul 2023•7 mos
Sabre • Bengaluru, India
  • ›Developed a Neo4j graph-based log analysis tool to visualize and inspect complex user flows in real time for product teams.
  • ›Led full-scale project migration from SVN to Git within weeks with zero workflow downtime.

Academic Credentials

Jan 2024 – May 2027 (Expected)

Master of Science in Computer Science (MS CS)

The University of Texas at Austin

Focus on distributed systems, middleware architectures, advanced computer systems engineering, and algorithmic scalability.

2019 – 2023

Bachelor of Technology in Computer Science & Engineering (B.Tech)

SRM Institute of Science and Technology

Specialization in systems programming, data structures, graph theory, and distributed computing.

Competitive Honors

2022

Winner — Smart India Hackathon (SIH)

Ministry of Education, Government of India

First prize nationally in India's premier nationwide engineering hackathon for building novel digital solutions.

2020

Runner Up — Hack36 2020

Motilal Nehru National Institute of Technology (MNNIT / NIT Allahabad)

Second place in a prestigious 36-hour national collegiate hackathon building high-impact software systems.

07 // INITIATE TRANSMISSION

Get in Touch

Have a distributed systems challenge, architecture inquiry, or high-scale backend role? Send an encrypted dispatch or reach out directly.

Direct Contact Endpoints

Primary Dispatch Email
Location: Bengaluru, India
Timezone: IST (UTC+5:30)
Response SLA: Within 24 hours

SES v3 Dispatch Engine

Rate Limited (3/10m)