What is aayi Companion? Inside Aayi.ai's 24/7 Maternal Health AI

aayi Companion (earlier known as AI Saheli — "saheli" means friend in Hindi) is Aayi.ai's 24/7 maternal health AI — built to be the medically reliable friend every expectant mother wishes she had at 2am when something feels wrong and she doesn't want to call her doctor.

The problem aayi Companion was built to solve

Every OB-GYN in active practice knows the pattern. A patient sends a WhatsApp at 11:30pm: "Doctor, I have a strange feeling in my stomach. Is this normal?" By the time the doctor reads and responds, it's midnight. The patient has already Googled the symptom and convinced herself she has preeclampsia.

Multiply this by 50–100 active patients and it's a significant drain on doctor time and wellbeing — while simultaneously not actually serving patients well, because the response time and anxiety gap are both problems.

aayi Companion was built to sit in that gap. Available instantly. Evidence-based. Calm.

How aayi Companion is different from a general health chatbot

The difference between aayi Companion and a general health AI (like asking ChatGPT about pregnancy symptoms) is significant:

• Clinical curation — aayi Companion's knowledge base is written, reviewed, and approved by qualified OB-GYN doctors and pediatricians. It's not trained on general internet content with medical content mixed in.
• Guideline alignment — Responses are aligned to WHO antenatal care guidelines, NMC protocols, FOGSI recommendations, and AAP pediatric standards — updated regularly.
• Context awareness — aayi Companion knows the patient's gestational age, medical history, and doctor-set parameters. A response about nausea at 8 weeks is different from the same question at 32 weeks.
• Doctor integration — When aayi Companion encounters a question or symptom it flags as potentially urgent, it doesn't just tell the patient to see a doctor. It notifies the treating doctor on their dashboard, with the conversation context.
• No hallucinations on clinical claims — aayi Companion is deliberately constrained to its curated knowledge base. It doesn't make up information. When it doesn't know, it says so and recommends contacting the clinic.

What aayi Companion can answer

aayi Companion covers questions across all three trimesters, postpartum, and newborn care:

• Common pregnancy symptoms — what's normal, what's worth monitoring, what needs immediate attention
• Nutrition and diet — foods to avoid, cravings, hydration guidance, meal planning basics
• Prenatal supplements — timing, dosage explanations, side effects, what to do when you miss a dose
• Exercise safety — what's safe at each trimester, when to stop
• Screening and test preparation — what each test involves, how to prepare, what results mean
• Labour signs — when to go to the hospital, what Braxton Hicks feel like vs. real contractions
• Postpartum questions — recovery, breastfeeding, baby care basics, postpartum mood changes
• Newborn and infant questions — feeding, sleep, fever, milestone development

What aayi Companion will not do

Equally important is what aayi Companion will never do:

• Diagnose a condition
• Recommend a specific medication or dosage change
• Override the treating doctor's instructions
• Provide false reassurance on symptoms that warrant clinical review
• Answer questions outside its curated knowledge base with invented information

The 90% midnight call reduction

Doctors using Aayi.ai report a dramatic reduction in after-hours contact within the first 30 days of patient onboarding — most citing reductions of 80–90%. The explanation is straightforward: the majority of midnight messages are anxiety-driven rather than clinically urgent. When patients have an immediate, reliable source of information, they use it instead of messaging the doctor.

The 10–20% of messages that remain are genuinely the ones that need the doctor — which is exactly the right filter to have.

Access to aayi Companion

aayi Companion is not a standalone product. It's part of the Aayi.ai platform, accessible to patients enrolled by a treating doctor. This design is deliberate — aayi Companion works best when it's connected to a clinical context, configured for the specific patient, and integrated with a doctor who reviews the conversation patterns.

Doctors interested in deploying aayi Companion for their patients can explore the platform through a free demo.

How aayi Companion Works: Step by Step

Understanding how aayi Companion actually functions — from onboarding to daily use — helps both clinicians and patients set accurate expectations. This is not a generic chatbot with a pregnancy skin. It is a structured clinical support tool built around a specific patient's gestational age, medical history, and care plan.

Step 1 — Doctor enrols the patient. The workflow begins at the clinic. The OB-GYN or their staff adds the patient to the aayi.ai platform with basic clinical details: name, phone number, estimated due date, and any relevant clinical flags (twin pregnancy, GDM, hypothyroidism, prior caesarean, etc.). This takes approximately 90 seconds. The patient receives a WhatsApp message with a download link and a one-tap registration process.

Step 2 — Pregnancy profile is configured. On first login, the patient completes a brief intake — her current trimester, any symptoms she's managing, dietary preferences, and language preference (English, Hindi, or Telugu). The AI companion uses this context to personalise every subsequent interaction. A patient at 8 weeks with gestational diabetes sees different default content than a patient at 32 weeks with a healthy singleton pregnancy.

Step 3 — Daily interaction begins. The aayi app delivers a daily "pregnancy moment" — a short piece of content relevant to the patient's exact gestational age. At 14 weeks, it might explain what is developing in the baby's nervous system this week. At 28 weeks, it explains why sleep position matters and how to manage it. This isn't generic week-by-week content; the system cross-references the patient's medical profile to surface relevant context.

Step 4 — Conversational queries via AI companion. When the patient has a question — at any hour — she types it into the aayi Companion chat. The companion draws on a curated clinical knowledge base reviewed by Dr. RVS Sai Sudha (OB-GYN) and Dr. Venkat (Pediatrician) to provide responses that are evidence-based, appropriately cautious, and genuinely useful. Questions about food safety, medication safety in pregnancy, symptom interpretation, and newborn care are answered in detail. The companion is explicitly designed not to provide diagnoses, not to replace clinical consultation, and to clearly flag when in-person assessment is needed.

Step 5 — Doctor reviews interaction patterns. The treating OB-GYN has access to a summary of the patient's aayi Companion interactions — not a verbatim transcript, but a structured view of topics she asked about, symptoms she reported, and compliance data. This creates a clinical feedback loop that most patient-facing tools don't provide: the doctor can see, before the consultation, that her patient has been asking repeatedly about pelvic pain, and come prepared.

Step 6 — Post-natal continuation. After delivery, the companion transitions from pregnancy support to newborn care support. Questions about breastfeeding, infant sleep, vaccination timing, and postpartum recovery are handled through the same interface. The companion then supports the pediatric vaccination schedule, creating continuity across the maternal and infant care journey.

What Questions aayi Companion Actually Answers

The most common question from clinicians evaluating aayi Companion is: what exactly does it answer, and how do I know the answers are safe? The knowledge base underlying aayi Companion has been curated and reviewed by practising OB-GYN and pediatric clinicians. It is updated periodically against current FOGSI guidelines, IAP recommendations, and WHO guidance. Here is a representative sample of the question categories it handles effectively:

Food safety in pregnancy: "Can I eat paneer during pregnancy?" "Is it safe to have papaya in the first trimester?" "What fish can I eat?" "Is it okay to have chai?" These are among the most frequently asked questions by pregnant women — often triggered by conflicting advice from family members — and they have clear, evidence-based answers that the companion provides accurately and with context.

Symptom interpretation: "I'm having sharp pains on my right side — what is it?" (likely round ligament pain at second trimester; companion explains what round ligament pain is, how to distinguish it from concerning pain, and when to see a doctor). "I haven't felt the baby move much today — is this normal at 24 weeks?" (companion explains what is normal at that gestational age, provides the kick count protocol, and clearly advises when to seek immediate care).

Supplement questions: "I'm feeling constipated — is it the iron tablet?" "Can I take paracetamol for a headache?" "My doctor said to take folic acid — how long do I need to continue?" These questions, if unanswered, lead to supplement non-compliance. The companion answers them in under 30 seconds, at any hour.

Labour and delivery preparation: "What does early labour feel like?" "How do I know if my water has broken?" "What should I pack in my hospital bag?" "When should I leave for the hospital?" These questions cluster in the third trimester and generate high anxiety when patients can't get answers quickly. The companion's responses are thorough and appropriately specific — including guidance on when to go to the hospital immediately.

Newborn and breastfeeding support: "My baby is feeding every hour — is that normal?" "I'm having pain while breastfeeding — what should I do?" "Baby has a rash on the face — is it normal?" "How do I know if my baby is getting enough milk?" The postpartum period has the highest density of anxious questions per unit time, and the companion's ability to provide calm, accurate responses is particularly valuable in the first four weeks after delivery.

Privacy, Safety, and Clinical Guardrails

For any AI tool operating in a clinical context, the questions of privacy and safety are not secondary — they are the central design challenge. Aayi Companion has been built with explicit guardrails that define what it will and will not do, and data handling practices that meet the requirements of India's Digital Personal Data Protection Act, 2023.

The companion does not diagnose. When a patient describes symptoms that could indicate a clinical condition, the companion provides context and reassurance where appropriate but explicitly does not name a diagnosis. It says: "What you're describing sounds like it could be related to [condition X], which is worth discussing with your doctor at your next appointment — or sooner if you notice [specific escalation signs]." This is the same response a knowledgeable OB-GYN nurse would provide, and it is clinically appropriate.

Emergency escalation is hardcoded. For a specific set of keywords and symptom patterns associated with obstetric emergencies — heavy bleeding, signs of severe pre-eclampsia, absent fetal movement, severe abdominal pain — the companion's response bypasses its standard conversational format and provides a direct, prominent directive to call emergency services or go to hospital immediately. This pathway cannot be customised or disabled by the clinic.

Data is stored in India, encrypted, and never sold. Conversation data between a patient and aayi Companion is stored on Indian servers, encrypted at rest and in transit. It is not used for advertising. It is not shared with third parties. The doctor who enrolled the patient has access to anonymised pattern data; they do not have verbatim transcripts of the patient's conversations. Patients can request deletion of their data at any time.

AI responses are not a substitute for clinical care. This is stated explicitly within the app on every session — not buried in terms of service, but surfaced at the start of each interaction: "I'm your AI pregnancy companion. I can provide information and support, but I'm not a substitute for your doctor's advice. For urgent concerns, always contact your doctor or emergency services." This boundary is reinforced, not concealed.