available for work

NYATHANI — SOFTWARE ENGINEER

HARIHARAN

python backend · prompt engineering · llm evaluation

dsabot.in — architecture sketch
async def log_practice(user, text):
    # Parse free text with regex + RapidFuzz
    entry = parse(text)
    log = await resolve(entry)
    await postgres.save(user, log)

# Separate workflows, not part of each log:
catalogs = leetcode_graphql + codeforces_rest
reviews = spaced_repetition + timezone_reminders
weekly_insights = optional_gemini
HYD, IN
17.38°N 78.48°E
EST. 2024
B.TECH CS

01 — WHO I AM

I build software that solves real problems. From a cross-platform DSA accountability platform spanning web, Android, and Discord to AI-powered evaluation workflows at Cognizant, my work combines Python backend engineering, prompt engineering, RAG, and LLM automation. I focus on building practical applications, connecting systems, and turning complex workflows into reliable software.

DSA Accountability
Platform

Web · Android · Discord Bot — dsabot.in

2026

Illustrative heatmap · Eligible LeetCode logs feed the spaced-repetition queue.

Python FastAPI React Flutter PostgreSQL / Supabase Discord.py
3 shared logging clients
2 problem catalog APIs
Weekly optional Gemini insights
librea — project outline

$ cat architecture.txt

├─ frontend: Next.js · TypeScript

├─ accounts & progress: Supabase

├─ authenticated EPUB streaming: Cloudflare R2

└─ offline reading: IndexedDB

Librea.

An installable EPUB reader with an approval-gated library, offline reading, and synced reading progress. View source on GitHub

July 2025 – Present

Prompt Engineer – Prompt Factory

Cognizant · Hyderabad, India

  • Built reusable components for the Prompt Factory auto-rater: policy-rule routing, RAG for policy context, and schema-defined JSON outputs for human review.
  • Consolidated about 320 policy checks into 129 atomic rules (roughly 60% fewer). Wrote few-shot prompts with explicit flag/no-flag criteria for shared and workflow-specific rules.
  • Automated concurrent LLM batch evaluation in Python. Reached 99% accuracy on selected categories against human-reviewed labels and refined prompts using per-rule error analysis.
  • Integrated multimodal Gemini into an auto-rater that evaluated 50,000+ assets across 12+ workflows, reducing manual review time. For YouTube ads, it compared original and enhanced video to flag clipped text, audio cuts, and blur.

MAY 2024 — AUG 2024

Full Stack Development Intern

Teachnook · Hyderabad

  • Built and deployed full-stack apps with Node.js, Express.js & MongoDB, with REST APIs, JWT authentication, and responsive interfaces.

2020 — 2024

B.Tech, Computer Science

Kakatiya Institute of Technology and Science · Warangal

  • Fundamentals in DSA, OOP, DBMS & systems — 150+ problems solved via Striver's A2Z sheet.
Python✦ FastAPI✦ React✦ Flutter✦ PostgreSQL✦ GraphQL✦ Gemini✦ SQL✦
REST APIs✦ JWT Auth✦ LLM Evaluation✦ Discord.py✦ Structured Outputs✦ MongoDB✦ RAG✦ Prompt Engineering✦
LET'S BUILD.