Mihir Vora

Software Engineer

Bengaluru, Karnataka, India2 yrs 10 mos experience
AI ML PractitionerAI Enabled

Key Highlights

  • Designed scalable microservices for AI-driven systems.
  • Built high-performance backend systems with millions of daily requests.
  • Led development of innovative AI-powered coaching solutions.
Stackforce AI infers this person is a Backend Engineer specializing in SaaS solutions with a focus on AI and data processing.

Contact

Skills

Core Skills

Backend DevelopmentMicroservicesData ProcessingAi DevelopmentCommunication SystemsWeb DevelopmentData Extraction

Other Skills

pythonDjangoCelerygRPCRedisDaskS3LLM AgentsOpenAI APITwilioOAuthDjango REST FrameworkScrapyBeautifulSoupRequests

About

Backend Software Engineer with 3.2+ years of experience building scalable microservices and AI-driven enterprise systems using **Golang and Python**. Currently working at **Convin.ai**, where I design and own high-performance backend systems powering AI agents, reporting platforms, and multi-tenant enterprise infrastructure. I specialize in building systems that operate at scale โ€” from processing millions of rows of data to serving millions of daily configuration requests โ€” while ensuring reliability, performance, and cost efficiency. ### ๐Ÿ”น What I Work On **AI Agent Tool Orchestration** Designed a modular, provider-agnostic orchestration framework enabling LLM-driven conversational agents to autonomously execute: * Real-time API calls * Calendar workflows * Mathematical computations * Knowledge retrieval using RAG Supports OpenAI & Google Gemini with clean abstraction layers. **Enterprise Asynchronous Reporting Platform** Architected a Celery-based distributed reporting system handling 200+ report types with: * Chunked pagination (50K-row batches) * Dask & Parquet-based intermediate storage * S3 large file generation & pre-signed delivery Built to reliably process datasets with millions of rows. **Centralized gRPC Configuration Service (Citadel)** Built a high-performance configuration service with: * Redis cache-first architecture * TTL jitter to prevent cache stampede * 2.5M+ daily requests handling * Automated tenant provisioning & migrations **AI-Powered Coaching & Sentiment Systems** Led end-to-end development of: * LLM-generated coaching insights from post-call transcripts * Mock call simulations using adaptive LLMs * Intelligent lead scoring (Goal Achievement, Interest Level, Qualification) * Event-driven webhook automation for sales workflows **SMS Communication System** Designed a Twilio-based SMS infrastructure including: * OAuth-secured webhooks * GSM-7 / Unicode segment calculation engine * Usage-based billing integration (Lago) for multi-tenant enterprises --- ### ๐Ÿ”น Tech Stack Programming Languages: Python, Golang, Java Frameworks & Libraries: Django, Flask, Django REST Framework, gRPC/Protobufs, Fast API, Restful API, Celery, Dask, Pytest, Swagger, Pydantic, Pandas, Numpy Databases & Messaging: PostgreSQL, MySQL, Redis, Parquet, RabbitMQ LLM & AI Tools: LLM Agents, LLM/Tool-Calling Integration Other: AWS, Docker, Kubernetes, Grafana, Microservices, distributed systems --- ๐Ÿ† First Prize โ€“ Convin Convo Nxt Hackathon (2024) Built an LLM-based dynamic coaching system that reduced 1.5+ hours of manual verification per account.

Experience

2 yrs 10 mos
Total Experience
2 yrs 10 mos
Average Tenure
2 yrs 10 mos
Current Experience

Convin

Software Engineer

Aug 2023 โ€“ Present ยท 2 yrs 10 mos ยท Bengaluru, Karnataka, India ยท On-site

  • โ—ฆ AI Agent Tool Orchestration: Designed and implemented a modular tool orchestration framework enabling
  • LLM-driven conversational agents to autonomously execute real-time actions including API calls, calendar workflows, mathematical computations, and knowledge retrieval via RAG (Retrieval-Augmented Generation). Supports multiple LLM providers (OpenAI, Google Gemini) with provider-agnostic architecture.
  • โ—ฆ Enterprise Asynchronous Reporting Service: Architected and implemented an enterprise-grade asynchronous
  • reporting service processing 200+ report types with Celery-based task queues. Engineered scalable data processing pipeline handling datasets of millions of rows through chunked pagination (50K-row batches), Dask DataFrames, and Parquet intermediate storage. Achieved reliable large-file generation with S3-based delivery using pre-signed URLs.
  • โ—ฆ Centralized gRPC Configuration Service (Citadel): Built a high-performance configuration service with Redis
  • caching layer serving tenant settings and feature flags to all platform microservices. Handles 2.5M+ daily requests
  • with cache-first architecture, TTL jitter to prevent cache stampede, and automated tenant provisioning (database
  • initialization, storage setup, schema migrations).
  • โ—ฆ Automated Agent Coaching System: Led end-to-end development of AI-powered training platform featuring:
  • โˆ— LLM-generated coaching responses and contextual examples derived from post-call transcripts to guide agent
  • improvement.
  • โˆ— Mock call simulations using adaptive LLMs for scenario-based performance assessment.
  • โˆ— Role-based content management and permissions system distributing training materials across LMS modules, forms, and feedback workflows.
  • โ—ฆ LLM-Powered Lead Metrics System: Built intelligent lead scoring system evaluating Goal Achievement, Interest
  • Level, and Lead Qualification using OpenAI/Gemini APIs. Implemented smart LLM call optimization with dedicated
  • calls for critical metrics and batch evaluation for cost efficiency.
pythonDjangoCelerygRPCRedisBackend Development+1

Simform

Software Developer

Jan 2023 โ€“ Jul 2023 ยท 6 mos ยท Ahmedabad, Gujarat, India ยท On-site

  • โ—ฆ Newton Scrapy - Web Scraping Pipeline: Built production-ready web scraping pipeline extracting structured data
  • and sitemaps from 100+ news websites using Scrapy framework, BeautifulSoup, and Requests library. Implemented robust error handling with retry logic, user-agent rotation, and rate-limiting mechanisms. Developed separate spider configurations to handle diverse website structures and formats, ensuring reliable data extraction across varying layouts.
  • Gained expertise in Python programming, covering topics such as functions, OOP Concept, asynchronous programming, multiprocessing, singleton design pattern, and advanced Python concepts.
  • Acquired knowledge of database concepts, including ACID properties, normalization, transactions, concurrency, and ER diagrams. Worked with MySQL, PostgreSQL, and MongoDB
  • Worked extensively with Django framework, mastering functions and class-based views, forms, models, templates, routing, Django Channels, signals, user authentication, mixins, permissions, and integrating Django
  • with DRF.
  • Gained proficiency in DRF, including serialization, deserialization, API views, permissions, versioning, class-based views, function-based views, viewsets, routers, nested serializers, throttling, token-based
  • authentication (JWT), CORS
  • Developed skills in integrating and working with Stripe payment gateway for payment integration.
Django REST FrameworkpythonScrapyWeb DevelopmentData Extraction

Education

Charotar University of Science & Technology (CHARUSAT)

Bachelor of Technology โ€” Information Technology

Nov 2020 โ€“ Nov 2023

Lukhdhirji Engineering (L.E.) College, Morbi

Jan 2017 โ€“ Jan 2020

Stackforce found 100+ more professionals with Backend Development & Microservices

Explore similar profiles based on matching skills and experience