Agentic Full Stack AI Engineer

AI Prof; Hyderabad; On-site; INR 279,000 - 335,000; Full time

Benefits offered by this job

Competitive stipend

Potential full-time role

Job summary

AI.Prof is recruiting a student with hands-on experience in Voice AI and multi-agent orchestration to build production-grade agents for hospital workflows.

You will work with Forward Deployed Engineers to deploy, evaluate, and monitor real-time AI agents, while learning advanced stacks like WebRTC, SIP/VoIP, and modern AI tooling in a healthcare context.

Qualifications

• Portfolio of non-trivial shipped AI/voice projects (GitHub, hackathons, live bots, or personal agents).
• Hands-on work with LLMs, Voice AI, or agentic frameworks is required.
• Genuine interest in healthcare as a high-stakes domain where software improves lives.

Responsibilities

• Deploy production Voice AI agents across telephony and web channels.
• Orchestrate production LLM pipelines with prompts, tool calls, RAG, and multi-step reasoning.

Skills

Agent orchestration

LLM integration

WebRTC & Voice AI

Backend APIs

Observability

Education

Final-year student, CGPA >7.5

Tools

LangGraph

CrewAI

AutoGen

LangSmith

Phoenix

Arize

LiveKit

Pinecone

Job description

About the Role

What You'll Do

• Deploy Voice AI Agents: Architect and ship production Voice AI agents that maintain natural, low-latency conversations across telephony and web channels under harsh, real-world network and audio conditions.
• Orchestrate Production LLM Pipelines: Build, test, and maintain robust LLM chains prompting, tool/function calling, RAG, and multi-step reasoning as deterministic production applications rather than one-off research scripts.
• Integrate Telephony Infrastructure: Connect SIP/VoIP, IVR systems, and telephony APIs directly with live AI voice agents for enterprise hospital deployment.
• Leverage Cutting-Edge AI Tools for Maximum Efficiency: Systematically identify, integrate, and evaluate new AI coding tools, modern agentic frameworks (LangChain, LangGraph, CrewAI, AutoGen), synthetic testing harnesses, and development utilities to continuously increase engineering speed, code quality, and system performance.
• Automate Healthcare Workflows: Own end-to-end automation for clinical workflows such as intake, scheduling, follow-ups, and documentation partnering directly with Forward Deployed Engineers (FDEs) to resolve real field friction.
• Instrument & Monitor Production Quality: Build observability into live agents to track voice quality, latency, task completion rates, and failure modes using operational data to continuously refine agent capabilities.
• Take Full End-to-End Ownership: Own your agents end-to-end: prototype them, write automated evaluations, deploy them, and ensure their long-term reliability once live.

What We're Looking For

Technical

• Agentic Orchestration & State Management: Practical experience with multi-agent frameworks and design patterns (e.g., LangGraph, CrewAI, AutoGen). Deep understanding of agent state persistence, human-in-the-loop (HITL) interruption loops, error recovery, and complex reasoning patterns (e.g., plan-and-solve, self-reflection).
• LLM Integration & Tool Ecosystems: Hands-on experience with commercial and open-source model APIs, advanced function/tool calling, and the Model Context Protocol (MCP) for secure tool exposure. Proficiency in building high-accuracy Retrieval-Augmented Generation (RAG) pipelines including advanced chunking, reranking, and hybrid search.
• Real-Time Voice AI & WebRTC Stacks: Deep familiarity with streaming audio pipelines, including Speech-to-Text (STT), Text-to-Speech (TTS), and low-latency audio streaming (e.g., LiveKit, Pipecat, Vapi). Practical understanding of WebRTC protocols, audio frame chunking, VAD (Voice Activity Detection), turn-taking logic, and interruption handling in live channels.
• Backend, Streaming & Data Infrastructure: Strong backend fundamentals using REST, gRPC, WebSockets, and webhooks. Experience managing structured data (SQL/NoSQL) alongside vector infrastructure (e.g., pgvector, Pinecone, Chroma) for semantic memory, caching, and fast retrieval.
• AI Tooling, Evals & Observability: Active utilization of modern AI-assisted development tools (e.g., Cursor, GitHub Copilot) for rapid prototyping and synthetic data generation. Experience implementing LLM observability, guardrails, and evaluation frameworks (e.g., LangSmith, Phoenix, Arize) to benchmark agent performance in production.

Good to Have

• Production WebRTC Deployment: Direct experience deploying an end-to-end, real-time WebRTC-based voice agent in a production environment (beyond a tutorial or demo build).
• Healthcare Interoperability: Prior exposure to healthcare data standards (HL7, FHIR) or deploying systems within regulated compliance environments (HIPAA, SOC 2).
• Real-Time Audio Pipelines: Familiarity with low-latency media servers (LiveKit, FreeSWITCH), WebSocket audio streaming, or noise suppression/VAD algorithms.

Mindset

• Extreme Ownership: You take ambiguous problems from initial concept to a shipped, reliable agent with minimal oversight.
• Bias to Action: You prioritize shipping a functional, testable agent today over endlessly polishing a theoretical architecture.
• Continuous Upskilling & Experimentation: You treat learning as a daily engineering habit. You actively test new tools, models, frameworks, and research papers before being asked, bringing back concrete findings to elevate team capability.
• Unsolved Space Mindset: Healthcare + Agentic AI has few established playbooks; you are energized by creating order out of ambiguity.
• Clear Communication: Ability to explain technical trade-offs, system limitations, and architectures clearly to hospital-facing and business teammates.

Qualifications

• Academic Standing: Final-year student with no active backlogs and a CGPA above 7.5 (or equivalent proof of exceptional analytical and problem-solving capability).
• Proven Execution: A portfolio of non-trivial, shipped projects (accessible via GitHub, hackathons, live bots, or personal agents) what you have actually built and shipped matters far more than formal coursework.
• Hands-On AI Experience: Demonstrable prior work with LLMs, Voice AI, or agentic frameworks (personal projects are valued equally with formal internship experience).
• Domain Drive: Genuine enthusiasm for healthcare as a high-stakes, high-impact domain where software directly transforms human lives.

Why This Role

Voice-driven, agentic AI in a regulated domain like healthcare is one of the hardest applied AI engineering problems today. While most teams building at this frontier are restricted to research labs, you will have the autonomy to ship real systems directly into partner hospitals within your first few months. You will get deep, unvarnished exposure to real-world deployment challenges with zero corporate bureaucracy.

What You'll Get

• Competitive Stipend: ₹25,000 – ₹30,000 / month (scaled directly to skill depth, execution velocity, and drive).
• FastTrack Career Growth: Top performers convert to a full-time role with a package of ₹6–10 LPA, based on impact and ownership shown during the internship.
• Direct Production Impact: Ship AI agents that run inside real hospital workflows on day one not sandbox projects.
• Field-Tested Engineering: Work shoulder to shoulder with Forward Deployed Engineers to put your agents through real hospital stress tests.
• Mastery of Advanced AI Stacks: Fasttrack your mastery across real-time WebRTC audio, multi-agent orchestration, telephony, and healthcare-grade production engineering.

Our Culture

This internship is not an ordinary employment placeholder. We are building the foundational infrastructure for healthcare AI, and we expect every team member to approach their work with extreme discipline, speed, and dedication. From day one, you are an architect of our engineering culture and a guardian of our technical standards.

Our Non-Negotiables

• Uncompromising Honesty: Speak with absolute clarity, especially when it is difficult or reveals a flaw.
• Hard Work & Humility: High technical execution capacity paired with zero ego.
• Impact for Humanity: Ensure every system built tangibly improves human health and hospital workflows.
• Relentless Excellence: High quality is our baseline; continuously raise the bar and obsess over the final 1% of detail.
• Extreme Ownership: Take total accountability. There is no "that is not my job" here.
• Bias to Action: Make sound technical decisions fast, execute fiercely, and adapt based on live production data.
• Solve Aggressively, Ask Fearlessly: Bring proposed fixes alongside problems, and raise technical blockers early to maintain momentum.
• Resourcefulness & Frugality: Embrace constraints creatively; leverage modern AI tools to maximize output efficiency.
• Disagree Openly, Commit Fully: Debate technical directions rigorously, then align 100% behind the selected path.

Note :

Equal Opportunity Employer: AI.Prof is proud to be an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees regardless of race, color, religion, sex, sexual orientation, gender identity, national origin, veteran, or disability status.