Backend Engineer

About the Role

Skit.ai is the leading multichannel AI platform for debt collection. Our Voice AI, Email AI, and SMS agents handle over 1 million calls a day across the US and India. We’re scaling fast, and our backend needs an engineer who thinks in throughput, fault tolerance, and latency budgets.

You’ll own the distributed systems that orchestrate parallel voice calls at massive scale, manage campaigns across millions of accounts, and keep our real-time voice AI pipeline humming under pressure. This isn’t glue-code work—it’s designing globally distributed systems where every millisecond and every dropped connection counts, requiring you to achieve massive horizontal scaling while strictly maintaining end-to-end low latency.

What You'll Own

• Voice AI at Scale: Design and operate the infrastructure that launches, monitors, and recovers thousands of concurrent voice sessions. You will manage parallel call orchestration, connection pooling, and real-time health tracking.
• Campaign Management Engine: Build systems that allow teams to configure, schedule, and throttle outbound campaigns across millions of accounts with complex retry logic and DNC (Do Not Call) compliance.
• Real-Time Pipeline Reliability: Own the availability and latency of the ASR → LLM → TTS pipeline. You will instrument, trace, and optimize the "hot path" to ensure calls stay within strict latency SLAs.
• Platform & Data Architecture: Evolve our backend services, queues, and data stores to handle 10x growth. Lead design reviews and make critical build-vs-buy decisions for core infrastructure.
• Integrations & APIs: Design robust APIs and webhook systems that connect our platform to client CRMs, payment gateways, and telephony providers with high throughput and reliable backoff semantics.
• Operational Excellence: Champion observability, incident response, and capacity planning. Build the dashboards and runbooks that ensure system stability during peak loads.

What You Bring

• Distributed Systems: Deep hands-on experience building and operating distributed backends, focusing on fault tolerance, consensus, partitioning, and high-availability failure modes. You should be adept at designing for consistency models (e.g., CAP theorem trade-offs), managing stateful services, and optimizing for latency-throughput trade-offs in high-volume production environments.
• Languages: Proficiency in Python, Go, or Java/Kotlin. You should have strong opinions on concurrency primitives and async I/O patterns.
• High-Throughput Infra: A proven track record with Apache Kafka, RabbitMQ, or SQS. Experience in stream processing, real-time pipelines, and event-driven architectures.
• Data Stores: Production experience with PostgreSQL, Redis, DynamoDB, or similar. You know when to reach for SQL vs. NoSQL.
• Cloud & Orchestration: Fluent in AWS or GCP. Comfortable with Kubernetes, Terraform, and CI/CD pipelines.
• Observability: Ability to build systems that are "debuggable by default" using distributed tracing and SLO-driven alerting (Datadog, Grafana, Prometheus).
• API Design: Experience scaling RESTful and/or gRPC APIs serving high-QPS traffic with proper versioning and auth.
• Ownership Mindset: 5+ years of building production systems. You are comfortable making architectural calls, leading RFCs, and mentoring junior engineers.

Bonus Points:

• Experience with Telephony / VoIP / SIP or real-time media streaming.
• Knowledge of LLM serving infrastructure or ASR/TTS pipelines.
• Experience in Fintech / Collections domains or TCPA/RBI compliance.
• Familiarity with WebSockets or WebRTC.