Kamailio (VOIP) Engineer

Job Title: AI/ML Engineer – Voice (2–3 Years)

Location: Bengaluru (On-site)

Employment Type: Full-time

About Impacto Digifin Technologies

Impacto Digifin Technologies enables enterprises to adopt digital transformation through intelligent, AI-powered solutions. Our platforms reduce manual work, improve accuracy, automate complex workflows, and ensure compliance—empowering organizations to operate with speed, clarity, and confidence.

We combine automation where it’s fastest with human oversight where it matters most. This hybrid approach ensures trust, reliability, and measurable efficiency across fintech and enterprise operations.

Role Overview

We are looking for an AI Engineer Voice with strong applied experience in machine learning, deep learning, NLP, GenAI, and full-stack voice AI systems.

This role requires someone who can design, build, deploy, and optimize end-to-end voice AI pipelines, including speech-to-text, text-to-speech, real-time streaming voice interactions, voice-enabled AI applications, and voice-to-LLM integrations.

You will work across core ML/DL systems, voice models, predictive analytics, banking-domain AI applications, and emerging AGI-aligned frameworks. The ideal candidate is an applied engineer with strong fundamentals, the ability to prototype quickly, and the maturity to contribute to R&D when needed.

This role is collaborative, cross-functional, and hands-on.

Key Responsibilities

Voice AI Engineering

• Build end-to-end voice AI systems, including STT, TTS, VAD, audio processing, and conversational voice pipelines.
• Implement real-time voice pipelines involving streaming interactions with LLMs and AI agents.
• Design and integrate voice calling workflows, bi-directional audio streaming, and voice-based user interactions.
• Develop voice-enabled applications, voice chat systems, and voice-to-AI integrations for enterprise workflows.
• Build and optimize audio preprocessing layers (noise reduction, segmentation, normalization)
• Implement voice understanding modules, speech intent extraction, and context tracking.

Machine Learning & Deep Learning

• Build, deploy, and optimize ML and DL models for prediction, classification, and automation use cases.
• Train and fine-tune neural networks for text, speech, and multimodal tasks.
• Build traditional ML systems where needed (statistical, rule-based, hybrid systems).
• Perform feature engineering, model evaluation, retraining, and continuous learning cycles.

NLP, LLMs & GenAI

• Implement NLP pipelines including tokenization, NER, intent, embeddings, and semantic classification.
• Work with LLM architectures for text + voice workflows
• Build GenAI-based workflows and integrate models into production systems.
• Implement RAG pipelines and agent-based systems for complex automation.

Fintech & Banking AI

• Work on AI-driven features related to banking, financial risk, compliance automation, fraud patterns, and customer intelligence.
• Understand fintech data structures and constraints while designing AI models.

Engineering, Deployment & Collaboration

• Deploy models on cloud or on-prem (AWS / Azure / GCP / internal infra).
• Build robust APIs and services for voice and ML-based functionalities.
• Collaborate with data engineers, backend developers, and business teams to deliver end-to-end AI solutions.
• Document systems and contribute to internal knowledge bases and R&D.

Security & Compliance

• Follow fundamental best practices for AI security, access control, and safe data handling.
• Awareness of financial compliance standards (plus, not mandatory).
• Follow internal guidelines on PII, audio data, and model privacy.

Primary Skills (Must-Have)

Core AI

• Machine Learning fundamentals
• Deep Learning architectures
• NLP pipelines and transformers
• LLM usage and integration
• GenAI development
• Voice AI (STT, TTS, VAD, real-time pipelines)
• Audio processing fundamentals
• Model building, tuning, and retraining
• RAG systems
• AI Agents (orchestration, multi-step reasoning)

Voice Engineering

• End-to-end voice application development
• Voice calling & telephony integration (framework-agnostic)
• Realtime STT ↔ LLM ↔ TTS interactive flows
• Voice chat system development
• Voice-to-AI model integration for automation

Fintech/Banking Awareness

• High-level understanding of fintech and banking AI use cases
• Data patterns in core banking analytics (advantageous)

Programming & Engineering

• Python (strong competency)
• Cloud deployment understanding (AWS/Azure/GCP)
• API development
• Data processing & pipeline creation

Secondary Skills (Good to Have)

• MLOps & CI/CD for ML systems
• Vector databases
• Prompt engineering
• Model monitoring & evaluation frameworks
• Microservices experience
• Basic UI integration understanding for voice/chat
• Research reading & benchmarking ability

Qualifications

• 2–3 years of practical experience in AI/ML/DL engineering.
• Bachelor’s/Master’s degree in CS, AI, Data Science, or related fields.
• Proven hands-on experience building ML/DL/voice pipelines.
• Experience in fintech or data-intensive domains preferred.

Soft Skills

• Clear communication and requirement understanding
• Curiosity and research mindset
• Self-driven problem solving
• Ability to collaborate cross-functionally
• Strong ownership and delivery discipline
• Ability to explain complex AI concepts simply

About the Role:

We are looking for an experienced VoIP Developer to design, develop, and maintain enterprise communication systems. You will work with Kamailio, OpenSIPS, Freeswitch, and Asterisk to build scalable VoIP solutions.

Skills Required:

• Hands-on experience with Kamailio, OpenSIPS, Freeswitch, Asterisk
• Knowledge of IMS & Diameter
• Proficiency in C / C++ programming
• Strong understanding of TCP/IP, LAN/WAN, VLAN/L3 VPNs
• Good knowledge of SIP/SDP, RTP/SRTP, TLS protocols
• Experience with MySQL / MongoDB
• Understanding of Enterprise PBX features
• Problem-solving and debugging skills
• Good written and verbal communication

Nice-to-Have:

• Building softPBX, Voicemail, ACD, IVRS applications
• Experience with Python, PHP, Java, JavaScript
• Linux administration and command-line skills
• Knowledge of WebRTC, firewall, security
• Familiarity with tools like cmake and git

What We Offer:

• Remote / flexible work options
• Opportunity to work on large-scale VoIP systems
• Collaborative environment with skilled peers

Job Opening: FreeSWITCH Developer

Location: Ecospace Business Park, Rajarhat, Newtown, Kolkata

Employment Type: Full-Time, Permanent

Shift: Day Shift

Experience: Minimum 4+ years in FreeSWITCH/VoIP Development

Responsibilities:

• Design, develop, deploy, troubleshoot, and maintain tools and services supporting our cloud telephony network.
• Customize FreeSWITCH for large-scale audio/video conferencing (1000–1500 concurrent calls).
• Expertise in SIP, RTP, RTCP, TURN, STUN, NAT, TLS.
• Hands-on experience with RTP Proxy and routed audio conferences.
• Understanding of SDP Protocol offer/answer mechanism.
• Work with load testing tools for FreeSWITCH audio conferences.
• Deploy and manage multiple FreeSWITCH instances using load balancers.
• Debug issues using packet captures (Wireshark/Ngrep).
• Collaborate with mobile/API teams for integration and support.
• Knowledge of codecs (PCMU, PCMA, G729, Opus) and open-source telephony technologies (FreeSWITCH, WebRTC).
• Familiarity with SIP servers (SER/OpenSER), proxy servers, SBC, SIPX is a plus.

Qualifications:

• Bachelor’s degree in Engineering (B.Tech/MCA).
• 4+ years of hands-on experience in FreeSWITCH development or related VoIP technologies.
• Strong knowledge of Linux environments and command-line tools.
• Basic to intermediate SQL knowledge.
• Proficiency in scripting (Bash, Python, Perl) for automation.
• Strong understanding of VoIP protocols (SIP, RTP), networking principles (TCP/IP, DNS, DHCP, routing protocols).
• Excellent troubleshooting skills with ability to resolve complex technical issues.
• Strong problem-solving, communication, and collaboration skills.

Industry Preference:

• Travel (Preferred)
• Any Industry (Required)