Skip to content

Applied AI engineer

SiftHubMumbai, Maharashtra, IndiaOn-Site

Estimated salary, not published by the employer: Est. ₹14–21 LPA

Needs a free Superjobs account. You'll finish on SiftHub's site.

Skill score

Summary

The Applied AI engineer owns SiftHub's retrieval, reasoning and generation pipelines end to end, from RAG architecture and prompt design to evaluation harnesses and agentic workflows. The role suits a production-minded ML engineer who wants to measure and steadily improve LLM output quality.

Function
Data Science
Estimated salary
Est. ₹14–21 LPA
Estimated for a mid-level role, ~6 yrs experience, India. Not published by the employer.
Work mode and location
On-Site, Mumbai, Maharashtra, India
Experience
6+ yrs
Employment type
Full-time

Job description

Overview

SiftHub's core value is the quality of the intelligence it surfaces to sales teams - delivered through a web app, Chrome extension, Microsoft add-in, and Google Sheets integration. That intelligence comes from a combination of retrieval, reasoning, and generation and it only works if every layer is built and evaluated with care. We need an Applied AI engineer who can design and own these pipelines end to end: not just get them working, but get them reliably good and measurably improving over time.

Roles and responsibilities

  • End-to-end RAG pipelines: retrieval architecture, chunking strategy, re-ranking, and prompt design
  • Evaluation frameworks: defining quality metrics, building eval harnesses, and tracking pipeline health over time
  • Agentic workflows and LLM integration: multi-step reasoning, tool use, orchestration, model selection, context management, latency, and cost optimisation
  • Collaboration with Backend engineers to serve NLP outputs at production quality and latency
  • Research-to-production translation: staying current and knowing what’s worth shipping

Skills required

Top skills

  • RAG pipelines
  • Evaluation frameworks
  • Python
  • LangChain

All skills (10)

  • LlamaIndex
  • FastAPI
  • OpenAI GPT-4
  • Anthropic Claude
  • Vector databases
  • AWS
  • Multi-agent orchestration
  • Enterprise data sources
  • LLM latency and cost optimisation
  • Sales tech and CRM data

Qualifications

  • Bachelor’s or Master’s degree in Computer Science or a related field, or equivalent practical experience
  • 3–6 years of production NLP/ML experience with shipped models or pipelines used by real users
  • Hands-on experience building and improving production-grade RAG systems
  • Experience defining quality metrics and building evaluation systems for LLM outputs
  • Solid Python engineering fundamentals with production-grade code (not just notebooks)
  • Experience with LangChain, LlamaIndex, FastAPI, or similar agentic frameworks
  • Experience working with LLMs such as OpenAI GPT-4 or Anthropic Claude
  • Familiarity with vector databases like Pinecone, ElasticSearch or pgvector
  • Experience deploying or working with cloud platforms such as AWS or Azure

Nice to have

  • Experience with multi-agent orchestration patterns
  • Familiarity with enterprise data sources (heterogeneous formats, noisy inputs, schema variation)
  • Exposure to latency and cost optimisation for LLM-powered products
  • Background in sales tech, CRM data, or B2B SaaS environments

Other details

  • Work on genuinely hard AI problems, not CRUD apps
  • Direct access to founder
  • Access to best-in-class dev tools and AI assistants to help you do your best work
  • Health insurance
  • In-office in Mumbai, we value building together in person

Company info

SiftHub is a deal orchestration platform that uses AI agents to automate RFP responses, sales collateral, deal briefs, and post-meeting follow-ups by sifting through connected tools such as Drive, Gong, CRM, and Slack. It is used by sales and solutions teams within revenue organizations to move deals forward.

More jobs at SiftHub