Location: Bengaluru, India (Hybrid) Job Type: Full-time

About Fibr.ai

Fibr is building the world’s first AI-Powered Agentic Workforce for marketing. We are replacing the old, manual experimentation stack with autonomous agents that personalize landing pages, run continuous A/B tests, and optimize conversion rates in real-time.

Backed by top-tier VCs (Accel) and founded by IIT/Stanford alumni, we are moving beyond simple "GenAI for copy" to Agentic RL systems that actively make decisions—allocating traffic, restructuring page layouts, and predicting user intent instantly. To fuel these agents, we need a robust, low-latency data engine.

The Role

We are looking for a hybrid Data & Machine Learning Engineer to build the backbone of our autonomous experimentation platform. You will not just be moving data; you will be creating the "nervous system" that allows our AI agents to observe, learn, and act.

You will own the end-to-end data lifecycle—from ingesting real-time clickstreams to serving real-time inference for our personalization models.

What You’ll Do

1. Build the Feedback Loops for Autonomous Agents

• Create Datasets for RL: You will design the state-action-reward data structures required to train our Contextual Bandit and Reinforcement Learning models. You will ensure our agents have clean, unbiased historical data to learn optimal strategies.

• Automatic Feedback Systems: You will engineer the "closing of the loop," ensuring that every user interaction (click, bounce, conversion) is immediately processed and fed back to the models to update their policies for the next visitor.

2. Own the Data Infrastructure

• Pipelines & Ingestion: You will build and maintain robust ETL/ELT pipelines to ingest high-volume data from diverse sources (Ad networks like Meta/Google, Client Clickstreams, CRM data) into our Data Warehouse.

• Data Warehouse Management: You will architect the schema and manage the warehouse (e.g., Snowflake, BigQuery, or ClickHouse) to support both analytical querying and ML training workloads.

• Data Quality & Governance: You will implement automated data quality checks (using tools like Great Expectations) to catch drift or corruption early, ensuring our agents are never learning from bad data.

3. Enable Real-Time Intelligence

• Scalable Inference: You will deploy ML models to production with strict latency constraints (targeting sub-50ms response times). You will optimize model serving infrastructure (using ONNX, TensorRT, or similar) to ensure personalization happens faster than the page load.

• Product Analytics APIs: You will maintain and optimize the high-throughput APIs that power our customer-facing analytics dashboards, ensuring clients can see the impact of our AI agents in real-time.

What We’re Looking For

• Experience: 6+ years of experience in Data Engineering or ML Engineering.

• The Stack:

  • Proficiency in Python (for ML/Airflow) and SQL (expert level).
  • Experience with Data Warehouses (Snowflake, BigQuery, or Redshift).
  • Experience with Workflow Orchestration (Airflow, Dagster, or Prefect).

• ML Ops: Experience serving models in production (Docker, Kubernetes, AWS SageMaker/Lambda). Familiarity with RL libraries (Ray Rllib, TF-Agents) is a massive plus.

• Real-Time Data: Experience with streaming technologies (Kafka, Kinesis, or Spark Streaming) is highly preferred.

• Mindset: You understand that an RL model is only as good as the data pipeline feeding it. You care about latency, uptime, and data correctness.

Why You’ll Love It Here

• Zero Legacy: We are building new paradigms in AI. You won't be maintaining 10-year-old code; you’ll be architecting the future of autonomous marketing.

• High Impact: Your work directly affects the "brain" of our product. Improvements in your data pipelines directly translate to higher conversion rates for our customers.

• Top Talent: Work alongside a high-density talent team from IITs, Stanford, and ex-unicorns (CRED, Disney, etc.).