Senior Full Stack Engineer - Agentic AI Systems
First seen by Alion on Sep 15, 2026.
Company : Soulpage IT Solutions Pvt. Ltd.
Location : Hyderabad - on-site
Experience : 5 - 9 years
About the Role :
- Soulpage is an AI-first technology company building enterprise AI products and platforms for clients including Fortune 500 and government organizations, alongside our own SaaS products.
- You will join a dedicated pod (engineers + BA + Delivery Lead + Solution Architect) working directly with the CTO office. This is product engineering with ownership - not ticket execution : you will take modules from architecture through production, working in the prod repositories to enterprise standards.
What You Will Do :
- Design and build backend services and APIs (Python) that productionise AI/ML and agentic workflows - from prototype-grade pipelines to reliable, observable, scalable services.
- Build and orchestrate agentic systems : multi-step LLM workflows, tool/function calling, retrieval pipelines, evaluation and guardrails.
- Work with graph and document data models - entity extraction outputs, knowledge-graph queries, large-scale document processing pipelines.
- Contribute across the stack where needed : React-based front ends for internal tools, review dashboards, and co-pilot interfaces.
- Own modules end to end : architecture proposals, milestone delivery with exit criteria, demos to all stakeholders, documentation as a first-class deliverable.
- Participate in fortnightly demos and architecture discussions with the engineering leadership.
Must-Have Skills :
- 5+ years of backend engineering with Python (FastAPI/Flask/Django), including REST/gRPC API design, async processing, and message queues.
- Hands-on production experience with LLM-based and agentic systems : at least one of LangChain/LangGraph, CrewAI, AutoGen, OpenAI/Anthropic SDKs with tool calling, or equivalent in-house frameworks - beyond tutorials : you have shipped and debugged these in production.
- RAG pipelines : chunking/embedding strategies, vector databases (pgvector, Pinecone, Qdrant, or similar), retrieval quality evaluation.
- Solid relational database skills (PostgreSQL) plus exposure to at least one NoSQL or graph store.
- Cloud-native development on AWS or Azure : containers (Docker), CI/CD, infrastructure basics; Kubernetes exposure preferred.
- Engineering rigor : automated testing, code review discipline, observability (logging/metrics/tracing).
- Strong written and spoken communication.
Good to Have :
- Neo4j or other graph database experience (strong plus - the platform's core is a knowledge graph).
- Document AI / intelligent document processing : OCR pipelines, layout parsing, entity and relationship extraction from technical documents.
- Full-stack experience with React/TypeScript.
- MCP (Model Context Protocol), agent evaluation frameworks, or LLM observability tooling (LangSmith, Langfuse).
- Domain exposure to energy, oil & gas, EPC, or industrial/manufacturing software.
- Experience working in AI-assisted development workflows (Copilot, Claude Code, Cursor) with strong code-review judgment.
What Makes This Role Different :
- Product ownership from day one - named module ownership with demos to client leadership, not anonymous sprint tickets.
- Frontier work : agentic AI applied to a hard, real domain where correctness matters, not a chatbot wrapper.
- Direct mentorship from Soulpage's Director of Engineering and Solution Architect embedded in the same pod.
- A clear growth path : this pod is phase one of a multi-phase programme expected to scale.
Hiring Process :
- Profile review + 20-min screening call.
- Technical deep-dive : backend design + agentic systems (90 min, includes a practical discussion of a system you've shipped).
- Problem-solving round with our Solution Architect.
- Culture and ownership conversation with the Director.
Skills
Full Stack, Agentic AI, Python, Flask, FastAPI, Django, RDBMS, Neo4j, Graph DBMS