Hi, I’m Abhinav. I build enterprise software end to end, from React front ends to Java and Python services.

I’m a software engineer at Cisco with 3+ years of taking features from design through production, focused on asynchronous processing, SQL performance, and LLM-powered retrieval.

Experience

  1. Software Engineer 2

    Aug 2023 – Present

    Cisco · Third-Party Software Compliance & Risk Management · Bengaluru, India

    • Built and maintain a FastAPI license-search service using BM25, BGE embeddings, Qdrant, and reciprocal rank fusion, finding ~3× as many correct license and document matches as the legacy system.
    • Reduced processing time by ~90% across compliance workflows with RabbitMQ-based asynchronous processing, batched database operations, and SQL query optimization.
    • Developed custom RAG infrastructure on PostgreSQL/pgvector with metadata-aware chunking and reranking, evaluated for retrieval quality, answer groundedness, and latency.
    • Led cleanup of ~32,000 duplicate components across four controlled phases, with zero downtime or data inconsistencies across dependent product releases.
    • Deliver full-stack features across React interfaces and Java/Spring Boot services; contribute to service designs and review pull requests across backend microservices.
    • Resolve Tier 2 production incidents using AWS, Kubernetes, and Kibana, working with business users from triage through resolution.

  2. Software Engineer Intern

    Jan 2023 – Jun 2023

    Cisco · Bengaluru, India

    • Expanded regression tests to cover over 95% of functional scenarios by repairing failing suites and adding missing cases.
    • Delivered backend features in Java/Spring Boot and frontend changes in React within an Agile team.

Projects

DeliveryLens

LLM engineering insights agent

An agent that correlates Jira and GitHub activity to flag delivery risks and write the status reports for you.

How it works

  1. IngestJira tickets and GitHub pull requests are upserted into PostgreSQL, with historical snapshots.
  2. DetectSeven delivery-risk patterns are checked, tracking review delays, status changes, and pull-request growth.
  3. SummarizeA local LLM writes summaries and triages stalled tickets, with numeric validation and template fallbacks.
  4. ReportDaily, weekly, and sprint reports are generated automatically.
  • Python
  • PostgreSQL
  • Ollama
Source: DeliveryLens (opens in a new tab)

PrepPilot

Interview preparation platform

A full-stack interview preparation platform with timed assessments, code execution, and progress tracking.

How it works

  1. Practice900 questions across 9 topics, with timed tests and adaptive spaced revision.
  2. QueueCode submissions enter a PostgreSQL job queue with expiring leases and retries.
  3. RunJava, C++, and Python solutions run in isolated Podman containers.
  4. ScoreHidden tests grade each run; passwordless OTP sign-in keeps progress private.
  • Java
  • Spring Boot
  • React
  • TypeScript
Live site: PrepPilot (opens in a new tab)Source: PrepPilot (opens in a new tab)

Skills

Languages

  • Java
  • Python
  • C++
  • JavaScript
  • TypeScript
  • SQL

Frontend

  • React
  • HTML/CSS

Backend & data

  • Spring Boot
  • FastAPI
  • REST APIs
  • RabbitMQ
  • PostgreSQL
  • JPA/Hibernate

AI & retrieval

  • RAG
  • BM25
  • Embeddings
  • Qdrant
  • pgvector
  • Ollama
  • LLM evaluation

Cloud & operations

  • AWS
  • Kubernetes
  • Docker/Podman
  • Jenkins
  • Linux
  • Kibana

Tools & testing

  • Git
  • JUnit
  • JMeter

Education & achievements

Education

Thapar University

B.E. in Electronics and Computer Engineering

Jul 2019 – Jun 2023 · Patiala, India

CGPA 8.5/10

Achievement

Cisco Hackathon

Built an LLM-assisted package source-discovery prototype that increased throughput by 30%, later integrated into existing services.

Contact

For roles, collaborations, or questions about my work, email is the quickest way to reach me.

abhinavsingla1111@gmail.com
View résuméDownload PDF