AI Systems
Model-backed workflows that still feel like dependable software.
I focus on prompt orchestration, backend guardrails, logging, caching, and safe delivery instead of demo-only AI.
AI Developer / Python Backend / Deployment / Analytics
Jithin Manilal
AI Developer • Python Backend Engineer
I build AI-enabled products, Python backend systems, and analytics workflows that are reliable in production and clear to operate.
- AI + backend implementation with Python
- Deployment, APIs, and data-heavy workflows
- NDA-safe client delivery and product thinking
Current direction
Shipping useful AI and backend systems with Python, databases, and production-ready delivery.
LLM-backed workflows
FastAPI + Django
Analytics automation
Deployment ownership
Python Backend
APIs, async processing, and database-backed services designed for clarity.
My strongest work is around FastAPI, Django, relational data models, integrations, and backend flows that teams can maintain.
Analytics Delivery
Turning operational data into usable outputs, reports, and internal tools.
I enjoy cleaning data, automating recurring analysis, and shipping internal systems that reduce manual work.
About
Building software that connects AI, backend engineering, and practical business workflows.
My work sits at the intersection of Python development, backend architecture, AI-assisted workflows, and deployment. I enjoy turning messy requirements into systems that feel structured, fast, and dependable.
I am especially comfortable with API design, database-backed products, scheduled processing, deployment setup, and the practical side of integrating AI into real user workflows.
What I enjoy most
Python services, AI integrations, dashboards, automation, and deployment workflows.
How I like to build
Clear APIs, structured models, observability, and maintainable systems over flashy complexity.
Skills
Depth where it matters: Python, APIs, AI integration, data, and production delivery.
Languages
2 toolsBackend Frameworks
3 toolsFrontend
2 skillsDatabases
3 skillsDevOps / Cloud
5 skillsCaching & Queues
3 skillsConcepts
5 skillsProjects
Selected work across AI features, backend systems, analytics, and deployment.
Client Project 01
Multi-Tenant AI Concierge
Confidential client system
A multi-tenant, retrieval-augmented AI chatbot platform that connects business knowledge, customer conversations, and human escalation workflows.
Overview
Built a location-scoped AI assistant platform for service businesses, managing knowledge documents, vector stores, bot configuration, and inbound conversations.
Challenge
Each tenant needed isolated knowledge and configuration while the assistant had to respond usefully, update the CRM, and hand conversations to people at the right time.
Solution
Implemented a Django backend around multi-tenant data models, OpenAI Assistants, retrieval workflows, webhook processing, intent classification, and CRM synchronization. Celery and Redis handled asynchronous work reliably.
Outcome
Created a scalable foundation for AI-assisted customer communication that kept business context, automation, and escalation paths organized.
Highlights
- Multi-tenant RAG and vector-store management
- OpenAI assistant and CRM workflow integration
- Async message processing and human escalation
Client Project 02
Agentic Chatbot Automation
Confidential client system
An OpenAI-powered chatbot platform that can take practical actions, including appointment scheduling, conversation management, and email delivery.
Overview
Developed an advanced chatbot management platform that turned conversational intent into operational actions through OpenAI tool calls.
Challenge
The assistant needed to do more than answer questions: it had to trigger dependable business actions without making the customer experience feel fragmented.
Solution
Connected the Django service to OpenAI Assistants and business APIs, enabling tool-driven actions such as marking conversations read, scheduling appointments, and sending follow-up information. React provided bot configuration controls.
Outcome
Delivered a configurable AI automation layer that connected conversational support to real business workflows.
Highlights
- Agentic tool calls for real workflow actions
- React configuration experience backed by Django APIs
- Celery-powered background execution
Client Project 03
Review Intelligence Platform
Confidential client system
A review-management platform that synchronizes business feedback, generates AI-assisted responses, and supports scheduled processing at scale.
Overview
Built API-driven review operations for onboarding locations, syncing customer feedback, producing AI-powered reply suggestions, and publishing responses to connected platforms.
Challenge
The system needed secure multi-service authentication, dependable background synchronization, and careful handling of user permissions and tokens.
Solution
Designed Django REST Framework APIs with OAuth, JWT, role-aware access, secure token handling, PostgreSQL data models, Swagger documentation, and Celery/Redis jobs for review processing.
Outcome
Provided a reliable backend for managing a growing review workflow while reducing manual response work.
Highlights
- AI-generated review response workflow
- OAuth, JWT, and role-based API access
- Scheduled synchronization with Celery and Redis
Client Project 04
Speed-to-Lead Analytics
Confidential client system
A data collection and analytics backend that transforms CRM webhooks and API activity into accurate, dashboard-ready lead response metrics.
Overview
Built a Django server that collected lead activity from CRM events and APIs, cleaned it, and stored normalized records for a custom analytics dashboard.
Challenge
Raw event data was incomplete and inconsistent, making it difficult to report a trustworthy measure of response time.
Solution
Implemented webhook ingestion, REST API synchronization, data cleanup rules, and PostgreSQL storage designed around the analytics questions the dashboard needed to answer.
Outcome
Turned fragmented operational events into a more dependable reporting dataset for monitoring lead response performance.
Highlights
- Webhook and API data ingestion
- Business-rule driven data cleanup
- Dashboard-ready PostgreSQL analytics model
Client Project 05
Mortgage Operations CRM
Confidential client system
A secure Django CRM backend for lead pipelines, loan workflows, reporting, team operations, and integrations across business systems.
Overview
Developed a backend for mortgage-brokerage operations covering lead management, loan applications, checklists, reporting, users, teams, and prospect onboarding.
Challenge
The platform required secure handling of sensitive workflows while coordinating data and notifications across CRM, storage, messaging, and reporting tools.
Solution
Built REST APIs, JWT and OTP authentication, encrypted payload handling, audit logging, background synchronization, secure file access, and integrations for communications, storage, and monitoring.
Outcome
Created an operational core that helped centralize complex workflows while retaining clear security and observability practices.
Highlights
- Secure APIs, encryption, and audit logging
- Celery/Redis workflow automation
- AWS file delivery and Prometheus/Grafana monitoring
Client Project 06
CRM Integration Suite
Confidential client systems
A set of Django middleware services that synchronize messaging, calls, calendars, and lender data between CRM and third-party platforms.
Overview
Delivered multiple integration services covering SMS, LINE messaging, RingCentral and Dialpad calling data, two-way calendar synchronization, and lender-directory onboarding.
Challenge
Every connected platform had different authentication, payload formats, event behavior, and reliability needs, yet users expected the CRM to remain the single source of truth.
Solution
Built focused Django services with API-key access, OAuth or SOAP token management, webhooks, scheduled sync jobs, and Celery processing for resilient cross-platform data exchange.
Outcome
Enabled business teams to work from one connected system while reducing duplicate updates and synchronization gaps.
Highlights
- Two-way calendar, messaging, and calling integrations
- OAuth, SOAP, webhooks, and API-key security
- Reliable asynchronous synchronization patterns
Experience
The shape of my work: product-minded backend delivery with growing AI and analytics depth.
Current Focus
AI-Enabled Backend Delivery
Python Services • Automation • Deployment
Building Python services that combine API design, AI-assisted workflows, analytics automation, and production-minded deployment patterns.
Project Strength
Backend Product Development
Django • FastAPI • PostgreSQL
Comfortable owning the full backend cycle: schema design, business logic, integrations, async workflows, and clean deployment handoff.
Working Style
Practical Engineering Mindset
Delivery • Maintainability • Clarity
I prefer systems that are easy to reason about, measurable in production, and built around useful outcomes rather than unnecessary complexity.
Contact
Let’s build AI-enabled products, APIs, or data workflows that are actually usable in production.
Open to AI developer, Python backend, platform, and analytics-focused roles
- Emailjithinmanilal@gmail.com
- LocationThrissur, Kerala, India
- GitHubView profile
- LinkedInConnect