Cybermind Works
Website:
cybermindworks.com
Job details:
AI Engineer (Full-Stack) — Imitate Labs
Location: Chennai (Thiruvanmiyur), On-site
Reports to: Founder & CEO (David)
Works alongside: Product Engineer (Sarathkumar)
Joining: September / October 2026
ABOUT THE ROLE
Imitate Labs is building Aurra Sales Agent, an AI-powered voice calling platform helping B2B businesses across India automate customer conversations. We serve Real Estate, Insurance, Automotive, Healthcare, EdTech, and NBFC clients.
We're looking for an AI Engineer whose primary focus is the AI layer itself — choosing and evaluating the models we run on, improving how our voice agents converse, bringing down inference and infrastructure cost, and building the intelligence products we sell to clients.
This is a builder's role in a small team, so the work doesn't stop at a recommendation. You'll write the Python services that power our voice agents and AI pipelines, build and extend the Node.js backend and Next.js interfaces around them, and deploy and operate what you ship on our cloud infrastructure.
KEY RESPONSIBILITIES
Core Focus — AI, Models & Intelligence
1. Model Evaluation & Cost Optimization
- Evaluate and compare LLM, Speech-to-Text, Text-to-Speech, and voice agent options to bring down inference and infrastructure cost without hurting conversation quality.
- Benchmark providers and model tiers on latency, accuracy, and cost-per-minute, and make a clear call on what we should be running.
- Own our cost-per-call and cost-per-minute numbers — token economics, prompt and context budgeting, caching, and understanding where the money actually goes.
- Assess self-hosted open-weight models against hosted APIs where the economics make sense.
- Build quick prototypes, test promising options against real call data, and take the winning one to production.
- Keep clear written notes on what was tested and why a choice was made, so decisions don't get revisited from scratch.
2. LLM & Conversational AI Development
- Build and improve conversational AI workflows for real-time voice agents.
- Integrate and work hands-on with OpenAI, Google Gemini, and Anthropic Claude APIs — streaming responses, tool/function calling, structured output, and context caching.
- Design prompt and context engineering strategies for spoken conversation, including multilingual and code-mixed Indian-language interactions.
- Develop agentic workflows with tool calling that reason across multiple business processes.
- Build test harnesses that measure response quality, latency, and cost per conversation, and improve all three over time.
3. AI Intelligence Platform
Build the next generation of Aurra's intelligence layer, sold as a premium product to clients:
- AI-powered lead scoring
- Conversation sentiment analysis
- Objection detection and pattern analysis
- Call quality and performance insights
- Personalized script recommendations
- Client health scoring
- ROI and cost-per-lead analytics
- Executive dashboards powered by AI-generated insights
You'll own the data pipelines that turn raw call data into these outputs, and the dashboards that present them.
4. AI Automation & Workflow Engineering
- Design and build AI-powered automation workflows using orchestration platforms (n8n, Make, or equivalent) and custom code where those platforms fall short.
- Automate hiring workflows, resume screening, AI interview assistants, and candidate evaluation pipelines.
- Build intelligent lead routing and follow-up automation.
- Integrate AI services with CRMs, WhatsApp, Email, dashboards, and internal applications.
- Build workflows with real monitoring, retries, logging, and failure recovery.
Also Part of the Role — Building & Running What You Ship
5. Backend & Systems Engineering
- Build Python services for the AI and voice layer — agent runtime, LLM and speech provider integrations, and data pipelines.
- Build and maintain our Node.js / TypeScript (NestJS) backend and REST APIs — campaign management, call orchestration, integrations, admin tooling, and client-facing endpoints.
- Work on real-time, low-latency systems: WebSocket services, streaming audio pipelines, event-driven flows, and sub-second response budgets.
- Design and evolve PostgreSQL schemas, write and manage migrations, and own data correctness.
- Integrate telephony providers, CRMs, messaging platforms, and client systems via REST APIs and webhooks.
- Write services that fail gracefully — retries, timeouts, idempotency, and structured logging.
6. Frontend & Product Interfaces
- Build and extend our Next.js / React / TypeScript applications: campaign dashboards, call and session logs, analytics views, and agent/prompt configuration tools.
- Turn AI capabilities into usable product surfaces — clean, functional interfaces you design the flow for and ship yourself.
- Implement real-time UI where the product needs it: live call state, streaming transcripts, and session timelines.
7. Deployment, Infrastructure & Production Ownership
- Deploy, operate, and monitor the services you build across staging and production.
- Own Docker builds, container orchestration, CI/CD pipelines, environment and secrets management, and the release process.
- Provision and maintain self-hosted Linux servers on cloud VMs running our platform.
- Run cloud infrastructure on AWS (compute, managed Postgres, object storage, networking, DNS/TLS).
- Set up monitoring, alerting, and logging; investigate and fix production issues in the systems you own.
- Manage cloud and inference cost — right-sizing, scaling, and eliminating waste.
- Share responsibility with the Product Engineer for overall platform stability and uptime.
8. Growth Engineering & AI Integrations
- Build AI-powered inbound lead processing and automated customer engagement flows.
- Integrate Meta Ads, websites, CRMs, and communication platforms.
- Enhance AI screening systems for both customers and job applicants.
- Support AI-driven experimentation for new features and revenue opportunities.
REQUIRED SKILLS & QUALIFICATIONS
- 1–3 years of hands-on software engineering experience shipping to production.
- LLM APIs in production — OpenAI, Gemini, Anthropic, or comparable — with real prompt engineering experience and a feel for where quality, latency, and cost trade off against each other.
- Comfortable comparing options with numbers — you can set up a fair test, measure latency, accuracy, and cost, and back your recommendation with data rather than impressions.
- Python — async/await and modern API frameworks (FastAPI or similar). Our AI, voice, and data pipeline work lives here.
- Node.js with TypeScript — building and maintaining backend APIs in a structured framework (NestJS, Express, or comparable).
- React and Next.js — you can build and ship a working dashboard end-to-end without a dedicated frontend engineer.
- PostgreSQL or a comparable relational database — schema design, non-trivial queries, and managing migrations (ORM experience welcome).
- Docker and Linux — you've built images, debugged containers, and worked comfortably on a Linux server over SSH.
- Cloud deployment (AWS, GCP, or Azure) — you have deployed, configured, and debugged your own services on real infrastructure, not just pushed to a managed PaaS.
- REST APIs, webhooks, and third-party integrations built and maintained in production.
- Git and collaborative development workflows.
- Practical experience with automation platforms (n8n, Make, Zapier, or equivalent).
- Ability to independently research, prototype, deploy, and operate solutions with minimal supervision.
- Strong problem-solving and experimentation mindset — comfortable with ambiguity and unproven ground.
PREFERRED QUALIFICATIONS
- Voice and speech technology: streaming Speech-to-Text, low-latency Text-to-Speech, voice activity detection, turn-taking and barge-in handling, or any voice-agent framework.
- Multilingual conversational AI for Indian languages — Tamil, Hindi, or English code-mixed speech.
- Experience building AI agents / agentic workflows with tool calling.
- Retrieval-Augmented Generation (RAG), vector databases, and embeddings.
- Real-time / streaming systems: WebSocket services, event-driven backends, and streaming media over the network. - Telephony: SIP, programmable voice APIs, call control webhooks, or contact-centre systems.
- Experience running or fine-tuning open-weight models on self-hosted infrastructure.
- Container orchestration (Kubernetes or a lightweight distribution) and CI/CD pipelines.
- Experience building analytics dashboards and reporting platforms.
- Observability tooling — structured logging, metrics, and tracing.
WHAT THIS ROLE IS NOT
- Not a research-only or notebook-only role, and not a pure application-development role either — the AI work is expected to reach production, and the full-stack work exists to put it there.
- Not a customer support or client-facing account management role.
- Not a design-heavy role — you'll build functional interfaces, not own visual design systems.
- Not a narrow specialist role — you'll move across AI, backend, frontend, and infrastructure as the work demands.
WHY JOIN IMITATE LABS?
- Build cutting-edge AI products used by businesses across India.
- Work directly with the Founder and influence product direction.
- Own high-impact initiatives end-to-end — ideation, build, deployment, and iteration.
- Experiment with the latest AI models and technologies.
- Ship products that generate measurable customer value and revenue.
- Join a fast-moving startup where innovation translates directly into business impact.
Note : Note: This position is with Imitate Labs. The recruitment and interview process for this role will be conducted exclusively by Imitate Labs.
Click on Apply to know more.
