I'm Ashwani Kharwar, an ai full stack engineer focused on backend-heavy product engineering, LLM orchestration, automation systems, and cloud-native delivery.
Portfolio
Ashwani Kharwar
lower AI inference cost
production engineering
LLM providers orchestrated
cloud-native deployments
About
Engineering where AI meets real business workflow.
Backend-focused full stack engineer with 2 years of production experience designing AI-powered systems, LLM orchestration pipelines, and cloud-native infrastructure. Built multi-agent architectures, workflow automation engines, and context engineering optimizations that cut AI inference costs by 50% and eliminated hallucinations in production. Owns DevOps end-to-end from Docker to Kubernetes migrations, Helm deployments, and high-throughput async processing at scale.
Capability Map
A practical stack for AI-native software.
The toolkit spans LLM systems, product backends, user interfaces, data stores, cloud delivery, and payments.
Languages
Backend Systems
Interfaces
AI Engineering
Data Layer
Cloud & DevOps
Commerce
Experience
Production ownership across AI, backend, and infrastructure.
- Designed and built a multi-agent orchestrator for Krut AI where specialized agents (planner, code generator, validator, deployer) autonomously hand off tasks based on context, enabling end-to-end app generation from a single user prompt.
- Applied context engineering across LLM pipelines by restructuring prompts, pruning context windows, and tuning token budgets - cutting AI inference costs by 50%, reducing hallucinations, and improving production response latency.
- Engineered a multi-LLM provider orchestrator routing requests across OpenAI, Anthropic, and Gemini with automatic failover on repeated provider failures, removing vendor lock-in.
- Led the full Docker Compose to Kubernetes migration: authored Helm charts, configured HPA, and added liveness/readiness probes for reproducible, scalable, self-healing deployments.
- Implemented shared NFS storage across Kubernetes pods for concurrent high-throughput async document workloads, eliminating data duplication across replicas.
- Designed an AI-driven candidate proctoring system with LLM-generated adaptive interviews, real-time multi-dimensional scoring, and automated structured assessment report generation.
- Engineered Accountant AI's workflow automation engine with pluggable nodes for email ingest, OCR extraction, LLM structured parsing, Excel export, and QuickBooks sync - replacing manual invoice data entry for finance teams.
- Built QuickBooks, bidirectional email, Celery Beat scheduler, and Razorpay payment flows for recurring workflows, accounting sync, subscriptions, real-time balance management, and automated billing.
Current operating mode
I work closest to the messy center: LLM reliability, async backends, finance workflows, deployment ergonomics, and the UI surfaces that make systems usable.
Selected Projects
Products with real surfaces, not just demos.
A few shipped builds across AI audio, education workflows, anonymous messaging, full-stack architecture, payments, storage, and production UI.
Audinexa (AI Audio Generation Platform)
Developed an AI-driven Text-to-Speech SaaS platform for generating, managing, and playing high-quality audio.
Built a custom Chatterbox TTS backend with voice cloning using FastAPI, for fast generation.
Integrated AWS S3 for secure audio storage, NextAuth for authentication, and Polar.sh for billing.
Designed a scalable full-stack architecture using Next.js, TRPC, Prisma, and PostgreSQL.
Study Notion (Ed-Tech)
Developed a full-stack EdTech platform enabling instructors to create and manage courses with role-based access.
Integrated video streaming, progress tracking, and cloud storage for seamless learning experiences.
Built secure payment workflows using Razorpay with dynamic coupon support and order history.
Implemented REST APIs, authentication middleware, and a modular architecture for scalability and maintainability.
Mystery Message (Anonymous Messaging Application)
Built a web app using Next.js for creating and decoding anonymous mystery messages.
Implemented secure backend APIs with Next.js API routes and MongoDB for message storage.
Designed a clean and responsive UI with TailwindCSS for smooth user experience.
Focused on privacy and anonymity, ensuring safe one-time message sharing.
Education
Foundation and trajectory.
Have an AI product, automation system, or backend that needs to become real?
Send the rough shape. I can help turn prompts, workflows, APIs, and infrastructure into something teams can use.